i

Proctored Versus Non-Proctored Testing: A Study for Online Classes

by

Danielle Truszkowski

Dissertation Proposal Submitted to the Doctoral Program

of the American College of Education

in partial fulfillment of the requirements for the degree of

DOCTOR OF EDUCATION

December 2019
 ii

Proctored Versus Non-Proctored Testing: A Study for Online Classes

by

Danielle Truszkowski

Approved by:

Dissertation Chair: Howard Moskowitz, Ph.D.

Committee Member: David Mapp, Jr., Ed.D.

Program Director: Elizabeth Johnson, Ed.D.

Assistant Provost: Crystal Neumann, D.B.A.

 iii

Copyright © 2019

Danielle Gallup Truszkowski

 iv

Abstract

Research-based consensus about the about the connection between proctored and non-proctored

assessments in online mathematics classes at the post-secondary education as related to success,

retention, and final exam scores, provides instructors and administrators with little guidance when

creating policies for online classes due to the lack of research. The purpose of the quantitative

research design, using a series of logistic generalized linear mixed models (GLMM) were

conducted to examine the relationships between proctor status and all the dependent variables,

final exam scores, success (passing the semester with a 70% or higher), and retention of online

mathematics students in the areas surrounding Baltimore, Maryland. Final exam scores, success,

and retention were compared between students who were given proctored versus non-proctored

tests during the semester to determine if there was a relationship between the test data in

relationship to proctored versus non-proctored tests in online mathematics classes. A relationship

status was determined by examining five different online mathematics classes with about 1900

students over a three-year period (2016-2018).

The results indicated students who took assessments during the semester in a proctored

class were more likely to fail the final exam. Proctoring semester assessments meant students

were 1.49 times more likely to fail the final exam. White and other races were more likely to pass

the final exam than African-Americans. With final grades, students who were proctored were less

likely to succeed. The result suggests students who attended the non-proctored assessments

during the semester were 1.53 times more likely to succeed in the course. Results predicting

retention demonstrated proctoring was not associated with retention. The only significant finding

showed Whites were more likely to be retained than African-Americans.

 v

Dedication

This dissertation is dedicated to my three daughters who have supported and encouraged

me throughout this journey. Each one of my girls has dealt with me working and researching

over the years during “their time.” I look forward to seeing Payton, Skylar, and McKenna

achieve professional and academic success, and make a difference in this world as we know

today.

I also dedicate this research to all the men and woman in online education. In this ever

changing world, more men and women are now able to follow their dreams and be successful

due to online courses and program offerings. I have created online courses, taught online

courses, and worked with people who were only able to be successful in the completion of their

degree due to the opportunities of online classes and programs.

 vi

Acknowledgements

I would like to acknowledge and thank my institution where I am employed for

contributing monetarily to a portion of the dissertation journey. Finances truly made a difference

in my journey and opportunities. I would also like to thank the staff for assisting me with data

collection. The process was completed in a timely manner with a considerable amount of time

and effort provided.

I would also like to thank Dr. H. Moskowitz for stepping up part way through the

process to be my chair, and promptly replying with feedback and answers while having a

positive attitude. My committee member, Dr. D. Mapp Jr. continued to provide solid feedback to

improve my research. Dr. T. Creighton provided feedback on my chapters and APA editing

suggestions.
 vii

Table of Contents

List of Tables ............................................................................................................................. xi

List of Figures ...........................................................................................................................xii

Chapter 1: Introduction .............................................................................................................. 1

Background of the Study ............................................................................................... .2

Statement of the Problem ............................................................................................... 4

Purpose of the Study ...................................................................................................... 5

Significance of the Study ............................................................................................... 6

Research Questions and Hypotheses ............................................................................. 8

Theoretical Framework .................................................................................................. 9

Definitions of Terms ...................................................................................................... 10

Assumptions ................................................................................................................... 12

Scope and Delimitations ................................................................................................ 12

Limitations ..................................................................................................................... 12

Chapter Summary .......................................................................................................... 14

Chapter 2: Literature Review ..................................................................................................... 15

Literature Search Strategy .............................................................................................. 18

Theoretical Framework .................................................................................................. 18

Online Courses................................................................................................................ 22

Success of Students in Online Courses .......................................................................... 26

Retention of Students in Online Courses........................................................................ 32

Proctored Tests and Non-Proctored Tests ……..............................................................35

Integrity in Online Courses ............................................................................................ 40

 viii

Chapter Summary .......................................................................................................... 45

Chapter 3: Research Method ...................................................................................................... 48

Research Questions and Hypotheses .............................................................................. 49

Research Design and Rationale....................................................................................... 50

Research Procedures.........................................................................................................51

Population and Sample Selection .....................................................................................52

Instrumentation................................................................................................................. 53

Data Collection .................................................................................................................54

Data Preparation ...............................................................................................................55

Data Analysis ...................................................................................................................56

Reliability and Validity ................................................................................................... 59

Ethical Procedures ............................................................................................................60

Chapter Summary ............................................................................................................ 61

Chapter 4: Research Findings and Data Analysis Results………………………………..……. 63

Data Collection ............................................................................................................. 64

Data Analysis and Results…………………………………………………………….... 65

Reliability and Validity ................................................................................................. 72

Chapter Summary ......................................................................................................... 72

Chapter 5: Discussion and Conclusion ....................................................................................... 76

Findings, Interpretations, and Conclusion………………………………………………..77

Limitations ................................................................................................................. 80

Recommendations .......................................................................................................... 82

Implications for Leadership ........................................................................................... 83

 ix

Conclusion………………………………………………………………………………..84

References .................................................................................................................................. 86

Appendix A: Consent Form ...................................................................................................... 108

 x

List of Tables

Table

1. Educational enrollment data for students in degree-granting 2-year higher education

institution in the United States .......................................................................................... ..3

2. Descriptive Statistics for Demographic Variables………………………………………..67

3. Descriptive Statistics for Independent and Dependent Variables ................................... ..69

4. Logistic Regression Statistics with Proctor and Demographics Predicting Final Exam…71

5. Logistic Regression Statistics with Proctor and Demographics Predicting Course Grade

for Success………………………………………………………………………………..72

6. Logistic Regression Statistics with Proctor and Demographics Predicting Retention…...73

 xi

List of Figures

Figure

1. Visual representation of how the principles of planned behavior relate to a student’s

characteristics in regard to ethical behavior in an online assessment………………20

 1

Chapter 1: Introduction

Online courses are increasing in popularity due to convenience and plasticity of not having

to go to a class (Gregory & Lampley, 2016). At a Mid-Atlantic community college (MACC), the

Mathematics Department aspires to work on raising the success and retention in online courses.

The mathematics faculty debated if proctored testing versus non-proctored testing is affecting

online students. With community colleges offering classes at a lower financial rate than four-year

colleges, there is an influx of part-time online students from the community, four-year colleges,

and military personnel who are stationed abroad. Community colleges are having the highest

growth rate in online courses in higher education (Ashby, Sadera, & McNary, 2011). Gender and

age difference can even make a difference when electing online classes (Haynie, 2015), which are

discussed in Chapter 2.

The study conducted was designed to contribute to the body of knowledge about the

connection between proctored and non-proctored assessments in online mathematics classes.

Chapter 1 introduces the quantitative research study where a series of logistic generalized linear

mixed models (GLMM) are conducted. The background of the study provides the research

situation, including a review of some literature relating to online courses and topics surrounding

proctored assessments in education. Chapter 1 presented the background of the problem, the

problem, purpose of the study, significance of the study, research questions and hypotheses,

theoretical framework, definition of terms used, assumptions, scope and delimitations, and

limitations to provide additional information about the study’s methods and procedures within the

literature.
 2

Background of the Study

The struggle by the mathematics department at a junior college located in the Mid-Atlantic

region of the United States, to implement rules about proctored and non-proctored testing may be

attributed to the mixed outcomes found in the literature research. Online institutions and

instructors have found strong evidence of a relationship between proctored and non-proctored

assessments. The final exam scores are a good indicator regarding the success and retention of

students. Research conducted on proctored and non-proctored testing has been done in different

parts of the country, with different content areas, at community colleges, and four-year colleges.

A number have shown high correlations with proctored and non-proctored assessments having

different outcomes (Nash, 2015), and others have shown to have approximately the same scores

(O’Connell, 2018).

When the first online degrees became available, computers were in more of an infancy

state, and government financial aid was not allowed to be applied. In the early 2000s pioneer

colleges of online education helped students’ complete degrees, and eventually, Federal financial

aid was being offered for online students (Ferrer, 2019). There is an estimate of four million

students who are taking online courses annually, and the number is expected to rise well into the

future (Ferrer, 2019). Although no one can put an exact number on online courses, there is no

telling where the courses are going, so institutions have to research the best methods for success

and retention in these now extremely popular classes. Despite the knowledge, the profession has

largely still concluded different results on the relationship between proctored versus non-

proctored assessments in relation to success and retention, due to mixed research. For instance,

Harmon and Lambrinos (2008) concluded in the study, cheating took place more often in non-

proctored assessments than proctored assessment which could affect success, while Watson and
 3

Sottile (2015) found there was no more evidence of cheating behaviors in the non-proctored

group than in a proctored setting, and Fask, Englander, and Wang (2104) determined the

difference was in the testing environment which produced a disadvantage for online student test

taking, which counterbalanced the advantage of cheating when the exam was non-proctored. For

example, the stress of going to an unfamiliar environment to test causes anxiety.

Online courses are increasing in popularity due to convenience and plasticity of not having

to go to a class (Gregory & Lampley, 2016). In fact, community colleges are having the highest

growth rate in online courses in higher education (Ashby, et al., 2011). In the decade prior to

2011, online course enrollment in community colleges increased 18.5% on average, while the

total U.S. enrollment in higher education increased 2.1% (Mitchel, 2017). The table below shows

the increases in two-year colleges in the United States. The growth of students taking online

courses, show the importance of studying online courses and instruction for improvement

(Mitchel, 2017).

Table 1

Educational Enrollment Data for Students in Degree-granting 2-year Higher Education

Institutions in the United States (U.S. Department of Education, 2014, 2015, 2016, 2017)

Total Percent taking Percent taking at least one Percent exclusively

any online course(s) online course, but not all taking online

of a student’s classes courses

Fall 2014 27.9 17.3 10.6

Fall 2015 28.8 17.6 11.2

Fall 2016 30.7 18.6 12.1

Fall 2017 32.0 19.2 12.8

 4

Statement of the Problem

The problem to be investigated was how students are failing online mathematics courses,

in which tests are proctored or non-proctored, at a higher rate than face-to-face courses. There

have not been standards put into place about proctoring assessments during the semester in online

mathematics courses, and the number of online classes the MACC is offering has increased. Since

learning and success of all students is what the school system is about, the problem is of great

importance. Perhaps more studies regarding online mathematics classes would provide further

insight as to how online mathematics courses can be taught in an electronic venue in order for the

academic achievement of students to be measured. Those impacted by the problem of students not

being as successful in online mathematics classes, are the students in the online mathematics

classes each semester, the instructors, those in charge of college finances, and community

businesses who hire students with an associate’s degrees.

The Mid-Atlantic community college serves over 62,000 students with about 45% of the

students being white, and 33% African American, 6% Hispanic or Latino, 6% Asian, 2%

multicultural, and 8% unknown. In addition, there are over 35% of students who are Pell Grants

recipients. Furthermore, the community college has about 29% of students who take an online

course, which has seen an increase each year.

The gap in the literature is between proctored and non-proctored assessments and the

relationship with success and retention in online mathematics classes specifically. Knowing if

proctoring assessments in online classes is affecting success and retention, and if so, what type of

proctoring and technology is best for the school and these classes, is important. The information

needs to be known for the financial view of the college and the success of the students. The gap

was closed by looking at multiple online mathematics classes and student test scores which have
 5

proctored assessments throughout the semester as well as those without proctored tests. Data were

collected from instructors on experiences with proctoring, the technology, and student responses

experienced.

A lack of understanding about the impact of proctoring assessments in online mathematics

education classes affects students, institutions, and communities. Non-proctored, online

assessments promise a low-cost opportunity to reach a large, heterogeneous, and geographically

diverse population (Gosling, Sandy, John, & Potter, 2010). In recent years, non-proctored, online

testing has become the main assessment mode in the academic field (Allen & Seaman, 2014). The

advantages of non-proctored testing can come at the cost of the lack of supervision, less

standardized test taking conditions, and less control over the student’s behavior. The question

may arise about dishonest behaviors in non-proctored assessments leading to unfair scores and

threats of the usefulness of online tests as a whole (Steger, Schroeders, & Gnambs, 2018). The

individual institution can be responsible to look into the needed information to make a

determination.

Purpose of the Study

The purpose of the quantitative correlation research design was to examine the

relationship of the following final exam grades (pass/fail), percent of success (passing the

semester with a 70% or higher), and percent of retention of online mathematics students. Final

exam scores, percent of student success, and percent of student retention were compared among

proctored versus non-proctored tests during the semester. The comparison determined if there was

a relationship in online mathematics classes. The proposed study was necessary for the MACC

and the mathematics department, to determine if the college can increase online success and

decrease dropout rates through the type of assessments given to students during the semester.
 6

Research indicates, online assessments are a challenge when determining a student’s performance

(Hollister & Berenson, 2009). Studies have found the percent of success in online classes are

lower than in the comparable face-to-face classes (Borzewski, 2016). If the research was not

conducted, there would not be a conclusion as to whether proctoring tests affect the success of

online students to meet or exceed those in face-to-face classes. The study contributes to the

knowledge base by determining if proctored assessments affect successful education in online

mathematics classes for students in at MACC.

Success is achieved for students by earning 70% or higher as the final grade in the course.

Cumulative final exam grades (pass/fail) were analyzed for students who had a proctored test

during the semester, and those who had not. Retention is upheld when students have passed a

mathematics course, signed up for the subsequent course needed in the student’s program of study

in order to earn a degree, certificate, or to transfer. Online student retention needs to be looked at

in each higher education institution as a wholistic, complete program approach in order to prepare

the students for a job after graduation (Kalinski, 2015). Data were sourced from secondary data

sources the Program Research and Evaluation Department has on file about students at MACC.

The study was designed to provide the Mathematics Department, online departments, and

administrators at MACC with information about how proctoring tests in online mathematics

classes are related to success, retention and final exam scores. The completed research study was

shared with the Mathematics Department, the online departments, and administrators of the

college.

Significance of the Study

This research study has been designed to be specifically applicable to instructors and

administrators. Administrators, instructors and students at MACC in online mathematics classes

 7

may benefit from research. Information was gained about possible effects of proctoring or not

proctoring assessments in online mathematics classes when determining success, retention, and

final exam scores. Decisions may be influenced by what percentage of assessments need to be

proctored to benefit the students, school, community, and ultimately attain a degree.

Alignment of the data shows how the items can all work together in order to achieve the

desired goals. Results of the data give information on research, and recommendations to

educational institutions, administrators, and instructors who are interested in a more in-depth

examination of online assessments in other geographic locations or cultural settings outside of the

study (Enago Academy, 2019). Failing to do the study may leave the community college

uninformed about the future of proctoring assessments in online mathematics classes while

valuing success and retention. Research designed to contribute to the body of knowledge in online

higher education as a predictor to effects of proctored and non-proctored assessments is of

importance.

Additional information with specific data focusing on the areas of success and retention of

students in online classes compared to traditional face-to-face students as well as gender and age.

Online success and retention remain a problem at the higher educational level which needs to be

addressed. For students, dropping out may typically mean unrealized possibility and lower

earnings over the course of a career, but the success of the university ends up being intertwined

with student success as well (Millea, Wills, Elder, & Molina, 2018). In addition to focusing on

success and retention, student integrity was considered, insuring credibility for students and the

institution. As online classes and programs have blown up in numbers, the academic integrity

issue has taken center stage with program design (Wagner, Enders, Pirie, & Thomas, 2016). The

study addressed issues important in online mathematics classes at the community college level.
 8

Research Questions

The following research questions guided the study based on the theoretical framework.

Research questions were developed to address the problem. To achieve the purpose of the study,

the following research questions were used for the quantitative study:

Research Question 1: What is the relationship between student final exam scores,

(pass/fail) for students with proctored versus non-proctored online math tests?

Research Question 2: What is the relationship between the percent of student success for

proctored versus non-proctored online math tests?

Research Question 3: What is the relationship between the percent of student retention for

proctored versus non-proctored online math tests?

Hypotheses

The following hypotheses were developed based on the research questions for the study.

Hypotheses were written based on the quantitative research design using a series of logistic

generalized linear mixed models for the statistics for analysis. To achieve the purpose of the study

the hypotheses were as follows:

H10: There is no relationship between final exam scores (pass/fail) of students with

proctored versus non-proctored online tests.

H1A: There is a relationship between final exam scores (pass/fail) of students with

proctored versus non-proctored online tests.

H20: There is no relationship between the percent of student success with proctored versus

non-proctored tests.

H2A: There is a relationship between the percent of student success with proctored versus

non-proctored tests.
 9

H30: There is no relationship between the percent of student retention with proctored

versus non-proctored tests.

H3A: There is a relationship between the percent of student retention with proctored versus

non-proctored tests.

Theoretical Framework

This research study was based on the theoretical framework of the Theory of Planned

Behavior, which was developed by psychologist Icek Ajzen. The Theory of Planned Behavior is a

cognitive theory which aims to predict and understand the relationship between human behavior

and motivation (Peters, & Templin, 2010). Based on the theory of belief, people use information

and reasoning to guide personal behavior. A key component to the model is behavioral intent,

which is influenced by the likelihood a behavior has an expected outcome as well as the risks and

benefits of the outcome (Boston University, 2018). Variables are used to predict an individual’s

behavioral intention, which in turn is used to predict actual behavior. Weight of an individual

variable may vary depending on the behavior and the population (Boslaugh, 2013). Researchers

have found a student’s likelihood of cheating depends on the degree to which the students can

rationalize cheating in a given circumstance (Eisenberg, 2004). Development of new behaviors

and practices is key when leaders are considering to what extent factors like motivation and

technology are impacting students cheating problem within each individual institution (Bolman &

Deal, 2008).

Research questions and hypotheses in the study were focused on the theoretical

framework in which proctored and non-proctored assessments were the independent variable.

Performance measures such as final exam scores, success, and retention were the dependent

variables. Chapter 2 offers evidence of relationships between online learning and assessments,
 10

success, retention, and possible integrity issues higher education institutions need to consider.

Additionally, Chapter 2 recaps research and information which exists within the same framework.

Definition of Terms

Definitions are provided below for the study’s dependent and independent variables.

Additional terms used in the research study having multiple meanings or may be ambiguous are

listed below as well. Definitions of terms for the study are:

Course ID. The name of the online mathematics course in the study (Community College

of Baltimore County, 2019).

Grade. Dependent Variable. The study had letter grades representing final exam scores. A:

90%-100%, B: 80%-89%, C: 70%-79%, D: 60%-69%, F: 0% - 59%. The grades are on the

student’s transcript for the class (Community College of Baltimore County, 2019).

Higher Education. Education past (after) the high school level (Kirkman, McNees, Stickl,

Banner, & Hewitt, 2016). Represented in the research and study as education at the community

college level.

Proctoring. Independent Variable. The study provided two different proctoring states,

which are represented by a dichotomous variable (“Proctored” or “Not Proctored”). The coding

used in the analysis has presence of proctoring or the absence of proctoring. A challenge to

consider with online classes is the perception associated with academic integrity, and

compromised testing due to undetected cheating with testing, which yields artificial high grades.

To address concerns, proctoring and proctoring software has been developed in hopes of

preventing academic dishonesty (Alessio, Malay, Maurer, Bailer, & Rubin, 2017).

Retention. Dependent Variable. The study provided two different retention states, which

are represented by a dichotomous variable (“Retained” or “Not Retained”). Retention continues to

 11

be an issue for colleges as the number of students taking an online course continues to increase.

Not only is the student who dropped affected, but so are other parties such as the college and

eventually the community (Cochran, Campbell, Baker, & Leeds, 2014).

Success. Dependent Variable. The study provided two different success states, which are

represented by a dichotomous variable (“Successful” or “Not Successful”). Success is when a

student earns a 70% or above as a final semester average for the given online mathematics class at

the Mid-Atlantic community college. As the popularity of online education continues to grow, so

do concerns about student success. Online students typically withdraw more often and receive

lower grades, compared to face-to-face students, and institutions need to have the goal of

improving online learning (Gregory & Lampley, 2016).

Term. The semester and year of the online mathematics course (Community College of

Baltimore County, 2019).

Assumptions

This study involved the collection of online mathematics student’s final exam scores,

percent of student success, and percent of student retention at the study site. The collection of data

were completed by the Program Research and Evaluation Department at the school. Online

mathematics classes used in the study are institutionalized, which means the students take the

same assessments during the semester, with the possibility of appropriate random numbers

inserted into each question. The assumption is, the instructors in each class told the truth about

whether proctored or non-proctored assessments were given during the semester. A questionnaire

was sent to each instructor in an email with each class in the study listed in a chart form.

Instructors checked the appropriate box with as to whether proctored or non-proctored

 12

assessments were given for each individual class. The assumption is necessary as there is no other

documentation about how the tests were given.

Scope and Delimitations

This study focused on a large community college in the Mid-Atlantic region of the east

coast. The sample consisted of about 1900 students who took institutionalized online mathematics

courses at the Mid-Atlantic community college. Standardizing online courses in the mathematics

department as well as final exams started in the fall of 2016. Due to evaluation and assessment

instruments across the same courses having multiple variables prior to 2016, the focus is restricted

to the study of the classes starting in 2016. Generalizing of the findings to other community

college online mathematics courses beyond the study site is likely limited, due to the scope of the

study.

A three-year period, from the fall of 2016 to the fall of 2018, was covered in the study. A

longer time period has not been selected due to time constraints associated with data collection.

MyOpenMath (2018) was the free online homework and assessment system used in all of the

classes in the study. All online classes were institutionalized and used by all instructors during the

three year period, so students were getting a similar experience.

Limitations

This study focused on success and retention as a measure of proctored and non-proctored

assessments throughout each semester. The given process is the only method of possible grade

differences which determine success, and ultimately possible retention differences and

relationships at the community college during the three-year period of 2016–2018. Low

completion rates are a problem for colleges who seek to serve students, employers who are reliant

on a workforce and economic prosperity (Levesque, 2018). A three-year period has been selected
 13

for the focused research to include a variety of instructors, both full time and adjunct, as well as a

variety and diverse population of students. During the time frame, the classes were

institutionalized, online, and included similar assessments, giving the students as close to the

same experience in each class as possible. These students were in five different levels of math to

help alleviate any potential student and instructor bias. More students were in the study due to

data which states students at community colleges during the given time, were more likely to sign

up for at least one online course than in four-year institutions (Lederman, 2018).

The study focused on proctored and non-proctored assessments as the independent

variable. Lurking variables having the potential to impact the student’s performance within the

research sample may have existed. Other demographic variables which could have a potential

impact on students’ performance include age, having to take care of family members, and the

number of hours students need to work outside of an education commitment (Glazier, 2016). The

research sample was diverse in these and students of all demographics in the study had an equal

opportunity to sign up for an online mathematics course in place of a traditional face-to-face

course.

Chapter Summary

Chapter 1 provided an introduction and overview of the research study. The purpose of the

quantitative study was to determine the degree to which proctored versus non-proctored

assessments were related to success and retention in an online mathematics class. Measuring the

data were accomplished by comparing the final exam scores (pass/fail), success and retention of

online mathematics students for a three-year period (fall 2016-fall 2018) in a large-sized

community college in the Mid-Atlantic region. The problem to be addressed was, there has been

minimal research-based consensus about the benefits of proctored assessments in online classes.
 14

There are benefits of researching the problem and relating student success and retention, leaving

community colleges with little guidance when establishing criteria for assessments in online

mathematics classes.

Chapter 1 introduced the study’s research questions, hypotheses, the theoretical

framework, and the methodology which was used to address these questions and hypotheses.

Chapter 1 included the research study’s definitions, assumptions, scope, delimitations, and

limitations. The study’s significance and importance are described and included contributions to

the existing body of knowledge relating proctored and non-proctored online testing to success and

retention with online mathematics students. Chapter 1 provided a background of literature

relevant to the research study. Chapter 2 provided a more in depth and comprehensive review of

the literature related to online classes, success, retention, proctored and non-proctored testing,

integrity, and the relationships compared amongst themselves and to face-to-face traditional

classes.
 15

Chapter 2: Literature Review

Online classes are becoming more popular amongst students across the United States, and

more colleges are offering online classes in significantly increasing numbers. The popularity of

online courses is partly due to convenience and plasticity (Gregory & Lampley, 2016). The Mid-

Atlantic community college (MACC), for example, increased online courses by 20% in just the

spring of 2018 (Community College of Baltimore County, 2018). Many students who enroll in

online courses have families and work responsibilities which make attending a class in person

difficult, in turn rendering online courses more appealing (Gregory & Lampley, 2016). Online

instruction is an area in which MACC needs to be prepared to make students as successful as

possible.

Neuman (2003) identified purposes of a literature review to be varying in capabilities of,

for example, finding gaps in knowledge on a topic, demonstrating a flow in the literature, and

showing the topic in a broader yet relevant context in order for the chosen topic to be credible.

The chosen topic determines whether proctored versus non-proctored assessments affect final

exam scores, success, and retention of online mathematics students. MACC is experiencing the

issue of students failing online mathematics courses at a higher rate than students who are

enrolled in face-to-face courses which are solely proctored. The background of the problem is, the

number of online classes MACC offers has been increasing, and the faculty does not necessarily

agree on the amount of proctoring for each online mathematics class. The problem is important,

considering students’ success and gained knowledge is the focus behind a school system (U.S.

Department of Education, 2015a). Those affected by the problem are the students enrolled in the

online mathematics classes each semester, the instructors, those in charge of college finances, and

community businesses who hire students with associate’s degrees.

 16

This study contributes to the literature with specific data in online mathematics classes.

The contributions are accomplished through observations of the online mathematics classes and

analysis of students’ final exam scores, success (passing with a 70% or higher), and retention. The

data were then be compared with those who received proctored assessments throughout the

semester, those who did not receive proctored tests, and those in traditional face-to-face classes.

About 29% of the total community college student population is enrolled in one or more online

courses—a percentage which increases each year. In the fall of 2018, 4% of the online students

were full time, while 96% were part time. While 67% of online students were female, 33% were

male. Regarding ethnicity, 6% were Hispanic or Latino, 43% were Caucasian, 39% were African

American, 7% were Asian, 4% were multi-racial, and 1% were unknown. 21% of the online

students were aged between fifteen and nineteen years, 47% between twenty and 29 years, 19%

between 30 and 39 years, 12% between 40 and 59 years, and 1% were aged 60 years or more

(Community College of Baltimore County, 2018).

The MACC has a student body of over 62,000 students who are eligible to take online

courses, while roughly 45% of the students are Caucasian, 33% are African American, 6% are

Hispanic or Latino, 6% are Asian, 2% are multicultural, and 8% are unknown (Community

College of Baltimore County, 2018). In addition, over 35% of students are Pell Grants recipients.

The total enrollment at MACC has been gradually decreasing over the past four years. In 2018

alone, enrollment decreased by 3% overall (Community College of Baltimore County, 2018).

The purpose of the quantitative research design was to determine the following variables

of online mathematics students at MACC: final exam scores, success, and retention. Quantitative

research was used to pose hypothesis constructs and frame the research questions (Waruingi,

2013). Correlational research attempts to quantify a relationship between combinations of

 17

variables for which analysis methods have been made available (Creswell, 2005). The study

determined whether there was a relationship in online mathematics classes between proctored

assessments, success, and retention. Retention occurs when students pass a mathematics course

and sign up for the subsequent course required in the program of study in order to earn a degree or

certificate or to be eligible for transfer.

A research study, as described, can contribute to the literature by determining whether the

proctoring of tests affects the success of online students by meeting or exceeding those in face-to-

face classes. Cumulative final exam scores were analyzed for students who have and have not

received proctored tests during the semester. These final exams are the same tests students in

face-to-face classes receive. Results of the study have been shared with the mathematics

department, the online departments, and administrators of the college.

The framework is based on the theory of planned behavior, which focuses on

understanding the relationship between human behavior and motivation (Boslaugh, 2013). The

theory includes variables believed to contribute a significant influence on outcomes related to

cheating behaviors in cases such as the taking of non-proctored assessments in online classes

(Madara, Namango, & Katana, 2016). The theory of planned behavior can be applied to the major

sections of the literature review, which discusses online classes, students’ success, students’

retention, proctored and non-proctored testing, and integrity in online classes. The data firstly

compares students who have received proctored exams with those who have not received

proctored exams during the semester, and secondly analyze how success and retention may be

related.
 18

Literature Search Strategy

Literature search strategies for the study were used continuously throughout the research

process. The main strategies involved searching online databases through EBSCO Library

Information Services from the American College of Education, searching the internet for

scholarly literature through Google Scholar, and visiting a library to search books and the online

system. The results returned reliable and credible databases, e-journals, magazines, and e-books,

all of which produced viable peer-reviewed literature. Research studies were identified by

searching key words such as: online classes, online courses, online education, online classes in

higher education, proctored versus non-proctored assessments, proctored versus non-proctored,

assessments in online classes, online assessments, success in online courses, success and

retention in online classes, retention in online courses, cheating in online assessments, and

integrity in online classes. Simple key word searches should not be expected to always lead to

optimal results listed in order of relevancy (Haigh, 2006). The library and librarians were

accustomed to locating research books on topics such as quantitative research, research methods,

data collection, and data analysis. Literature search strategies can somewhat easily provide

knowledge of what research has been conducted, what results were found, how the information

may be of assistance, and what still needs to be completed (University of Leeds, 2018).

Theoretical Framework

The theoretical framework for the research study is based on the theory of planned

behavior, which was developed by psychologist Icek Ajzen. The theory of planned behavior is a

cognitive theory aimed to predict and understand the relationship between human behavior and

motivation (Peters & Templin, 2010). The theory is based on the belief, people use information

and reasoning to guide behavior. A key component to the model is behavioral intent, which is
 19

influenced by the likelihood a behavior has an expected outcome as well as the risks and benefits

of the outcome (Boston University, 2018). Variables are used to predict an individual’s behavioral

intention, which in turn are used to predict the individual’s actual behavior. The weight of an

individual variable may vary depending on the behavior and the corresponding population

(Boslaugh, 2013). Figure 1 provides a visual representation of how the principles of planned

behavior relate to a student’s characteristics in regard to ethical behavior in an online assessment.

Students'
Attitude: Students'
Preceived Behavior
opportunity to Beliefs
cheat/not cheat

Subjective
Normative Norm: Students' Intention: Behavior:
Beliefs: friends/family Students intend Academic
Students want would want to be ethical and misconduct/
to be ethical in them to be not cheat Academic
education ethical in
education honesty

Students'
Precieved
Control Beliefs Behavioral
Control/Lack of
Control

Figure 1. Left-to-right view of planned behavior concerning students and ethical behavior in

regards to cheating. Students’ attitudes, normative beliefs, and controlled beliefs come together.

Empirical researchers have confirmed the validity of the theory of planned behavior in

regard to predicting human behaviors (Chu, Chen, & Sung, 2016). Per the theory, there are three

predictors of human behavior. First, there is an attitude toward behavior, which is the extent of
 20

positive or negative appraisal a student would possess regarding a given behavior. Second, there

are subjective norms, which are a student’s perceived social expectations or pressure to conform

to the given behavior. The third predictor is perceived behavioral control, which is a student’s

perceived ability to carry out the behavior based on past experience (Yang, Choi, & Lee, 2018).

The theory of planned behavior includes demographic variables believed to contribute to a

significant influence on outcomes related to cheating behaviors in situations such as the taking of

non-proctored assessments (Madara et al., 2016). Some of these behaviors are factors of gender,

discipline, and education level (Whitley & Keith-Spiegel, 2002). Some concepts students

typically manifest within themselves to justify behavior are: “It was not my fault,” “It was not a

big-deal,” “You were just as bad when you were my age,” and “My friends needed me. What was

I going to do?” (McQuillan & Zito, 2011). The concepts listed present justification as to why

students may violate ethical codes which would otherwise be supported (McQuillan & Zito,

2011).

Bolman and Deal (2008) declare most effective leadership in theories such as the theory of

planned behavior establish meaning and predictability in what is considered a disordered world.

Researchers have found, a student’s likelihood of cheating depends upon the degree to which the

student can rationalize cheating in a given circumstance (Eisenberg, 2004). Curtis, et al. (2018)

have determined both self-control and perceived behavioral control, when contributing to the

prediction of plagiarism and cheating, are the best fit in the event which direct paths from

perceived norms to plagiarism behavior are specified. The information suggests schools which set

strong anti-plagiarism norms, such as honor codes, and pursue the enhancement of students’ self-

control may reduce student engagement in plagiarism and cheating. The need for the development

of new behaviors and practices is key when leaders are considering to what extent factors such as
 21

motivation and technology affect the student cheating problem within the institutions (Bolman &

Deal, 2008).

There are several limitations and undetermined factors of the theory of planned behavior

in regard to integrity and cheating. The theory of planned behavior assumes, regardless of the

intentions, the involved students have attained opportunities and resources with which may be

used to become successful while practicing the desired behavior. While the theory does consider

normative influences, the theory does not account for all environmental or economic factors,

which may influence a student’s intention to perform a behavior (Boston University, 2018). In the

future, researchers may want to investigate additional factors of theory, such as the role of

religion. Religion may increase the variance explained in any model of cheating (Al-Dossary,

2017). Finally, research found men were more likely to cheat than women, which may lead to

varying behaviors of men versus women (Al-Dossary, 2017).

The research questions and hypotheses in the study were focused on the theoretical framework

in which proctored and non-proctored assessments were the independent variable. Performance

measures such as final exam scores, success, and retention were the dependent variables. Chapter

2 offers evidence of relationships between online learning in regard to assessments, success,

retention, and possible integrity issues, higher education institutions should consider. Research

and information which exist within the same framework is reviewed in Chapter 2.

Research Literature Review

Instructors and administrators in education throughout the United States must make

informed decisions when creating and establishing criteria for online mathematics courses. The

problem is, there is minimal research-based consensus about proctored and non-proctored testing.
 22

The literature review examined some of the debated topics of online courses, like integrity,

retention, and proctoring.

Online Courses

As online course offerings evolve, more strategies have been researched for the purpose of

improving the online learning experience for students and instructors alike. Woods and Bliss

(2016) believe increasing students’ engagement in online discussions and the assessment process,

providing feedback, and overcoming some challenges in the facilitation of online course

discussions are all the best practices for online education. Reflective assessments and grading

rubrics are recommended for students’ learning, as well (Woods & Bliss, 2016). Martin, Wang,

and Sadaf (2018) verified, instructors need to establish presence, connection, and engagement in

order to achieve the most effective learning. The more thoroughly the students are involved, the

more strongly the students want to remain in the class in order to learn in future classes (Martin et

al., 2018).

Reddy and Andrade (2010) recommended course rubrics for assignments and assessments,

along with a timely instructor grading system included for students’ satisfaction in online courses

as well. With these recommendations, not only may students’ learning and engagement increase,

but the exceptional instructors using these methods could become mentors who may help enhance

other teachers’ instruction and online classrooms (Reddy & Andrade, 2010). Student capabilities

in online classes are different than those needed to succeed in face-to-face classes, and instructors

need to determine what increases students’ satisfaction, self-motivation, and the ability to manage

time wisely in order to become successful, independent learners based on the course requirements

(Lee, Stringer, & Jianjun, 2017).

 23

Jayaratne and Moore (2017) conducted a research study at North Carolina State University

to determine students’ perceptions toward online classes. With about 90% of students possessing

experience taking an online class, a student’s decision to take an online course or a face-to-face

course seemed to be based primarily on class scheduling. Platt, Raile, and Yu (2014) concluded,

students in higher education settings felt there were fewer opportunities to interact in online

courses with instructors and classmates, which made students believe slightly less knowledge was

gained. Despite this, students claim online courses possess the same rigor as face-to-face courses.

The most helpful online instruction methods students believed may help close the gap, according

to Jayaratne and Moore (2017), were instructional videos, PowerPoints with recorded narratives,

video recordings of face-to-face courses, instructional audios, and reading materials. The

information encourages instructors to find ways to replace visual and verbal instruction used in

face-to-face classes. The materials may assist older students who feel online courses are more

difficult in general (Jayaratne & Moore, 2017).

Beck and Milligan (2014) indicated positive effects on a student’s persistence and success

have been recognized for quite some time in face-to-face programs, and more studies need to be

conducted with online students. Institutions need to investigate students’ family backgrounds,

reasons for attending online courses, demographics, and experiences in order to improve online

learners’ institutional commitment and success (Beck & Milligan, 2014). Ouzts (2006) claimed

online courses designed to promote a sense of community may address many of these concerns.

Ouzts (2006) suggested, concerning online courses and the continuous, rapid growth, there

is still considerable concern among educators in regard to the quality of the learning experience.

Students themselves have expressed concerns regarding the lack of contact with faculty and peers,

which has made the students sense isolation in cyberspace as well. These feelings of isolation and
 24

disconnect may contribute to negative learning outcomes and lower retention. Regardless,

students are choosing online courses because of the flexibility, convenience, and accessibility

(Sanford, Ross, Rosenbloom, & Singer, 2017).

Distance learning models have evolved over the past few decades. Social media and

technologies, according to Friedman and Friedman (2013), encompass a wide variety of learning

strategies which can enhance online learning, such as blogs, wikis, online social networking, and

virtual worlds. Knowing many students are not retained and quit college because courses were

uninteresting, requires educators to more actively motivate success. Over 80% of chief academic

officers believe massive open online courses (MOOCs) are an important factor in helping teachers

learn about online pedagogy and improve online classes, according to Allen and Seaman (2014).

Online education is being used as an alternative to the violent schools to which some students are

assigned, and thus becoming important for institutions to improve online classes when many

disadvantaged children are being taught online and need help achieving success (Friedman &

Friedman, 2013).

Peslak, Kovalchick, Wang, and Kovacs (2018) studied students enrolled in computer

information systems courses at three colleges during the 2016–2017 academic year. The students

comprehensively preferred the face-to-face course delivery method over the online method. Sole

online course delivery was considered moderately effective according to the surveyed students.

As reported by Kovacs, Peslak, Kovalchick, Wang, and Davis (2017), when asked which option

students would select if given the choice between online or face-to-face presentation, 54% of

students preferred the latter, while 46% preferred the former. Hybrid courses were found to be

effective by 84% of students. Contrary to solely online courses, neither age nor gender differences
 25

were found to be significant regarding the effectiveness of hybrid courses. The fact of online

courses being perceived as less favorable, is a call for improvement in online delivery methods.

Various cultures may perform differently in online classes, as well. Tucker (2014)

identified conditions within and outside online class environments which supported academic

success among male college students of color. Some of these factors within online environments

support academic success, and are convenience and flexibility, a colorblind environment, faculty

support and immediate interactions, and institutional support (Tucker, 2014). Beyond the online

environment, factors such as self-efficacy and educational resilience affect online classes. In

addition, Moore (2014) noticed some African American students withdrew from online courses

and specified the decisions were associated with inadequate technology, computer skills, and

support as well as perceived course difficulty.

Students of color preferred in-person social supports, which may lessen the feelings of

isolation and alienation often experienced by students in online classes (Tucker, 2014). Economic

factors may influence academic underachievement among African American male college

students. According to the National Center for Educational Statistics (2015), 50% of African

American male students are raised by single mothers. Single-parent families in which African

American males are usually raised are typically associated with a greater incidence of poverty

(United States Census Bureau, 2015). Greater incidences of poverty and lower incomes are

associated with subordinate educational outcomes (Child Trends Databank, 2015). Travers (2016)

believed African American male students as well as others stemming from lower socioeconomic

populations with lower standards in grade school need to keep up with employment obligations

while remaining motivated in school. Due to the outside factors affecting these students, the
 26

pursuance of higher education online is not only appealing, but necessary for the goal of a degree

or certificate.

In the United States, there have been some challenges faced in online courses. For

instance, various institutions considered different levels of online classes as being “online”.

Online courses have now been defined as having at least 80% of the course material delivered

online (Allen, Seaman, Babson Survey Research Group, & Quahog Research Group, 2015).

Online and distance education studies have faced challenges due to the content being reported and

not reported, especially in regard to blended, hybrid, and fully online courses. The content is

something leadership may be able to develop more affluently while providing additional

innovations for long-term success. According to Allen and Seaman (2014), another issue involves

comparing retention in online courses to those of face-to-face courses. Due to there being online

students who may have never attended a college class if there were exclusively traditional classes,

the comparisons may not match. If students drop out of an online course due to the environment,

the dropout may be a reflection not of the class, but of the student’s nature.

Success of Students in Online Classes

Success is the goal of education. Studies have found success in online classes are lower

than those in the comparable face-to-face classes (Borzewski, 2016). Researchers have delved

into demographic issues, academic concepts, enrollment, and environmental factors surrounding

success. Online courses offer convenience and flexibility, although many of these students who

want to sign up for online courses possess characteristics which may lead to high risk of academic

failure (Gregory & Lampley, 2016). Not only was success in online classes at the community

college level lower than those in face-to-face classes, there were significant differences in success

rates based on the types of instructional methods used (Gregory & Lampley, 2016). Borzewski
 27

(2016) explains how higher education professionals need to be educated on instructional methods

and potential impacts of students in online classes in order to achieve the greatest success.

Shotwell and Apigian (2015) surveyed students to determine potential impacts which

affect online students. The impacts covered student insight, academic resources, and core

motivating factors which lead to success in class. For instance, students mainly used homework as

the first tool to learn a new topic. Instructors should be aware, and tailor approaches toward

student learning and behaviors which are associated with success in regard to online testing

(Shotwell & Apigian, 2015). Matika (2012) concluded, in an online study of mathematics courses

at a community college, the difference in the lower success rates and retention rates in these

courses compared to face-to-face courses may be a result of student self-efficacy, communication,

and presentation on the instructors’ part. The findings both conclude the instructors can make a

genuine impact and a worthy difference on student success.

Research reveals students who learn in online classrooms may be affected by self-efficacy

and success academically, however a higher success can be reached than with students in

traditional classrooms who do not utilize technology (Ozerbas & Erdogan, 2016). The

incorporation of digital-based activities can increase students’ motivation and result in higher

success rates. Many students can relate to technology and be comfortable communicating through

it, which can promote connections between everyday life and the courses themselves (Ozerbas &

Erdogan, 2016). Success and technology in online classes are becoming even more important with

the expansion of online classes, even among on-campus and typically traditional students

(Murphy & Stewart, 2017). All online student success among community college students and

four-year college students cannot be generalized in all cases. There may be too many variables,

such as the student diversity, or type of assessments given, which vary within each type of college
 28

and should be considered, and thus individual colleges may wish to conduct research individually

(Ashby et al., 2011).

If higher education continues to grow with online offerings, there should be an expansion

of support for students and instructors alike. Rey (2010) described how the online modality of

teaching is gaining ground at such a rapid rate, but without guidelines for quality and merit, there

are too many unknown factors for students and instructors. Betts, et al., (2013) focused on student

success in online classes and explained how accessibility is one support which should be a

priority when designing a course for the purpose of increasing success. Accessibility and online

courses are a collective responsibility which requires a commitment from the institution and

instructors alike. Many institutions have used online education as a way to reduce costs and

increase enrollment, which students have bought into for alleged ease and convenience (Rey,

2010). Institutions should maintain student and instructor support as a priority for future success.

Although many success rates have been lowered in online courses than in comparable

face-to-face classes, clarity is still needed as to whether other variables such as student

demographics and history are related to the issue as have been in some remedial math classes in

higher education (Borzewski, 2016). Friedman and Friedman (2013) suggested the issues in

online education are raising the standard and making the classes interesting for a diverse group of

students—a feat instructors have tried to accomplish in face-to-face classes over the years. Hybrid

classes, according to the study, may be the best method to teach courses for college students while

maintaining interest and connection. The business world is implementing like strategies in stores

such as Wal-Mart and Target. Wal-Mart and Target sell an abundance of products both in store

and online, and both options possess particular benefits. Friedman and Friedman (2013) believed

online learning, to some degree, should be taught as early as kindergarten in order for students to
 29

understand the valuable tools of discipline and persistence, which could ultimately help students

acquire a degree, and in order for institutions to understand how many variables are related.

California community colleges, according to Johnson, Cuellar Mejia, and Cook (2015),

believed in giving instructors new pedagogical tools and ways to track students in order to

implement a more data-driven, integrated, and systematic approach geared toward improving

online learning. Online education has allowed for increases in enrollment without the requirement

of additional classroom space. The setting of standards and provision of tools for instructors

contributes toward the increased availability of online courses to those who could not reach a

brick-and-mortar classroom—with just as much success—and may close the gap in success for all

students. Bernard, et al. (2014) determined there is a strong association between the strength of

communication and success in online courses compared to face-to-face classes. Teaching

standards, provisions of tools, and strong communication skills to instructors, may increase

cognitive engagement and may be promoted by strengthening interactions and success in online

courses.

California community college students enrolled in online developmental and traditional

courses over a four-year period were studied by White (2013) as well. The study sought to

determine whether gender had an impact on success. Overall, women were found to be more

successful than men in any modality of course taking (White, 2013). There existed no difference

in success in online classes versus face-to-face classes. Gregory and Lampley (2016) agreed,

women outperform men on average, but went a step further. The study looked at non-Pell Grant-

eligible students and nontraditionally aged students in online classes. The study determined these

students had higher success rates, even though there was typically more work-related, financial,

and personal responsibilities and burdens. Another point noted was, students who are not Pell
 30

Grant eligible receive alternative financial aid and are only able to continue receiving financial aid

if a certain GPA is maintained as well as a full-time student status, which often determines if

students can stay in college or not.

Shea and Bidjerano (2014) analyzed data to compare degree completion rates of

community college students enrolled in online courses during the first year enrolled, to students

enrolled in all face-to-face courses the first year. The conclusion of the study noted, students who

registered for online classes during the first year of college had higher rates of attaining a degree

than those who did not take online courses during the first year. Researchers often agree, the most

successful students who earn degrees in online courses are self-disciplined, goal-oriented,

responsible, and organized (Johnson & Berge, 2012). These students are abreast in time

management, multitasking, and critical thinking, and tend to take responsibility for learning, and

are capable of working independently. These are typically characteristics of an adult learner, who

is often considered a nontraditional student. The students are usually more mature and possess

prior knowledge and life experiences to which can be related to education and have effectively

introduced these elements back into online education classes (Johnson & Berge, 2012).

A data analysis of online students and face-to-face students was conducted with more than

twelve years’ worth of studies, and Means, Toyama, Murphy, Bakia, and Janes (2009) determined

online students generally performed better than face-to-face students. Understanding how students

succeed in online courses may lead to even far greater success. Such an understanding may

significantly contribute to initiatives which could double the number of American college

graduates. Effective practices for engaging and motivating students to persist toward success can

be most effectively tackled through collaboration with other online learning organizations (Moore

& Fetzner, 2009). Together, instructors and institutions can achieve more in less time.
 31

Quality education is a goal for all means of instruction, the Online Learning Consortium

(2015) explained, and included satisfaction, fairness, and rigor for students, as well as instructor

and peer interaction. Effective professors help students achieve learning outcomes and require the

support of a technical infrastructure and training in online skills in order to achieve these goals.

Although learning effectiveness is focused on ensuring online students are provided with a high-

quality education which is at least equivalent to traditional students, there is no implication online

learning experiences should duplicate what occurs within traditional classrooms (Online Learning

Consortium, 2015). Xu and Jaggars (2014) suggested, although gaps persevere between online

and traditional classes in persistence and course grades, in order to ensure positive and effective

online learning experiences, the traditionally underserved students should be truly supported in

each case. If these students are not, the continued expansion of online learning may strengthen

rather than diminish educational inequity (Xu & Jaggars, 2014).

Shea and Bidjerano (2014) suggested students do not achieve degrees or certificates in

online programs at the community college level when enrolling during the first year. Contrary to

preexisting expectations, the study additionally noted, students with certain background

characteristics who take some online classes early on do have a better chance of obtaining

credentials as opposed to solely traditional peers. Institutions should find importance in

determining the characteristics of the student population, as the proportion of a diverse group of

higher education students taking one or more online course is at an all-time high, with online

enrollment growing more than five times faster than the total enrollment (Bailey, Barton, &

Mullen, 2014). In fact, 16% of higher education students are now registering and learning

primarily through online classes (Bailey et al., 2014).

 32

In order to direct students to the correct classes, community colleges may consider

requiring the completion of an assessment prior to enrollment (Xu & Jaggars, 2011). The

assessment may be tailored to each student by directing the student to a specific college session if

claiming basic technology skills. Furthermore, students may be advised to look into face-to-face

courses if the student scores poorly on the assessment, or the student may be asked to schedule an

appointment with an advisor if those classes do not fit availability wise. Western Governors

University, which is solely based online, has adopted the strategy of trying to provide a

personalized education by allowing students to forego formal courses and directing the student to

take only the courses required for the degrees the student is trying to earn, which leads to the

possibility of earlier graduation (Western Governors University, 2018).

Retention of Students in Online Classes

Success and retention can go hand-in-hand with education. Elam (2013) discussed some of

the issues faced in online classes and introduced ways to increase retention. For instance, the

study determined, demographics often make a difference in retention, and an orientation class for

a school’s online classes improves retention and success. Students who are educated on how these

classes work, operate, and are organized, possess the ability to improve retention and success

(Elam, 2013). James, Swan, and Daston (2016) added, although learning outcomes in online

courses are similar to those in face-to-face courses, retention and success are larger issues in

online classes and should be addressed. Students who took online classes early on at the

community college were significantly less likely to return to school for subsequent terms, while

those who took a high proportion of online credits were significantly less likely to graduate or

earn a certificate (James et al., 2016)

 33

Since student retention is an issue in higher education, Black (2018) conducted a study

which provided educators strategies to implement and improve retention rates. The study

determined how critical and pivotal instructors working toward improving students’ attendance

and success in order to improve retention can be. Retention strategies found by Cochran, et al.,

(2014) for online classes include student engagement activities, learning communities, learning-

centered environments, and readily available information on student services. Engagement in the

first year of college and the method by which tests are conducted online are crucial for student

retention. Cochran et al. (2014) have identified prior student performance (GPA) and class

standing (freshman, sophomore, junior, senior) as important characteristics associated with

student retention in online courses, which may be used to identify at-risk students early on in the

coursework process. When a student drops a class, the student loses time, money, and the self-

confidence necessary to complete the program. Multiple parties are affected when a student drops

a course and fails to return to a class or school in general.

Research, proven strategies, and theoretical models have been implemented to increase

student retention (Gomez, 2013). Obtaining information on students’ backgrounds, past academic

achievements, socioeconomic statuses, and personal characteristics are habitually collected by

institutions and may be used for research purposes. The readily available student data from a

specific institution can be an effective predictor for student retention. The information can provide

administrators and educators with what is necessary to create strategies and models which inspire

and motivate degree and/or certificate completion. These strategies and models can help students

overcome weaknesses and encourage greater persistence and retention (Gomez, 2013). The

integration of supports and strategies for students and instructors alike in online courses not only
 34

improves retention rates but enhances the role community colleges play as a whole (Travers,

2016).

To increase retention in online courses, models need to be continuously reviewed, such as

models in modern literature. The review allows for the understanding of changing behaviors in

online students and faculty. Institutions can then determine how student characteristics fit together

in an online educational unit and how retention and success can be upheld namely as online

enrollment rises (Papia, 2016).

In Virginia community colleges, almost half of the students are enrolled in an online

course, and the data were used for a study conducted by Jaggars and Xu (2010) at Columbia

University. Regardless of preparation, online students were more likely to withdraw or fail than

face-to-face counterparts. Some students reported frustration with typing skills, course navigation,

and the learning management system. To empower students, the study suggests, online courses be

explicitly designed around the unique content of the online class rather than by the pedagogy and

materials of the comparable face-to-face class. Educators need to remove the barriers of time and

space which are experienced in the traditional educational system by offering uniquely designed

courses at learners’ doorsteps (Gul, Shafiq, Mahajan, Shafi, & Shah, 2018).

An additional item which can affect issues relative to retention rates, according to

Richardson (2018), is the lack of financial resources at historically black colleges and universities,

which primarily affects online classes. The research cites, these institutions serve a positive

societal purpose as well as an instrumental role in the education of lower socioeconomic classes

which other colleges do not offer. Components are able to be preserved of the African American

ethnic identity in both online and face-to-face classes. Salvo, Shelton, and Welch (2017) believe
 35

online education can possibly offer a color-free environment in which students are less likely to

be judged by race and more likely to be treated equally.

According to Moore (2014), African Americans as a whole are more likely than other

demographic groups to take entire undergraduate programs online. Ironically, historically black

colleges and universities are typically slower to create online programs for students than are other

universities. In fact, only 18% of historically black colleges offer online degrees (Flowers, White,

Raynor, & Bhattacharya, 2012). Certain cultures and socioeconomic classes such as the African

American community can build connections, support one another, and improve retention rates

(Blue, 2018). Other important concepts to consider in regard to online classes, African

Americans, and the improvement of retention and dropout rates are mentors’ responsibilities and

higher education preparation for student success. Mentoring and building connections are actions

considered so important, and potentially even save money can be saved for students in regard to

retention rates (Blue, 2018).

Proctored Tests and Non-Proctored Tests in Online Classes

The use of online exams as part of the evaluation process has been studied and researched

to some extent. In the study conducted by Ardid, Gomez-Teiedor, Meseguer-Duenas, Riera, and

Vidaurre (2015), students were given online proctored exams, online non-proctored exams, and

an online training homework task. The analysis reveals students’ online grades depended on the

way the online exam was administered. For instance, in the non-proctored setting, the results were

biased toward higher ratings, while a greater dispersion of results was observed in the

proctored environment. The study noted, non-proctored assessments can have security and

honesty concerns. Advances in technology, according to Weiner and Hurtz (2017), have spurred

innovations in the security of assessments and the delivery process. Online testing has become
 36

increasingly sophisticated even with an absence of published research. Results of the study

conducted by Weiner and Hurtz (2017) supported the equivalence of kiosk-based, remote online

proctored exams and proctored onsite assessments in test centers. With the given research, a

community college may choose to implement the cheaper option within the institution.

Non-proctored online assessments promise a low-cost opportunity to reach a large,

heterogeneous, and geographically diverse population (Gosling, et al., 2010). The pessimistic

viewer might conclude, some participants cheat every time if given the opportunity, regardless of

countermeasures or expected consequences (Diedenhofen & Musch, 2017). Those with a more

positive outlook believe students do not cheat if the opportunity is not given. Whatever the case

may be, administrators and instructors of non-proctored assessments are encouraged to adopt

strategies to identify potential cheaters with information on reaction times or non-reactive

behavioral data (Diedenhofen & Musch, 2017).

Since non-proctored online assessments are often compromised by the lack of control over

students’ test-taking behavior, Steger, et al., (2018) examined mean score differences and

correlations between proctored and non-proctored assessments. Perceived consequences of the

assessments, countermeasures against cheating, and the susceptibility to cheat were considered.

Standardized mean differences indicated higher scores in non-proctored assessments. The results

demonstrated, non-proctored assessments are biased by cheating. Some studies found no

statistical differences between proctored and non-proctored assessments (Ihme et al., 2009), while

others reported much higher scores for non-proctored tests or, infrequently, for proctored tests

(Carstairs & Myors, 2009).

Varying results were reported in regard to how often students were cheating, which ranged

from below 2.5% to 7.0% (Tendeiro, Meijer, Schakel, & Maij-de Meij, 2013). In an online
 37

survey, 25% of students reported, cheating on a problem may be most appropriate when the

problem is difficult to search online (Jensen & Thomsen, 2014). A reason for the varied results

may be the diverse educational settings in which non-proctored assessments are often

administered (Allen & Seaman, 2014). No significant effects were found with other variables,

which suggests the score differences between proctored and non-proctored assessments are not

affected by students’ expected consequences of assessment results. In cases when solutions to

problems were not easily found on the internet, mean score differences were roughly zero (Allen

& Seaman, 2014).

The findings validate previous research, suggesting, some tasks are more susceptible to

cheating than are others (Diedenhofen & Musch, 2017). For example, Bloemers, Oud, and van

Dam (2016) examined cheating strategies for various online assessments. Cheating was

determined to be most often present for problems which could be tampered with through online

searches. Cheating did not affect problems in online non-proctored tests for which complex

reasoning was required (Bloemers et al., 2016).

Recent technological developments have changed the way researchers collect data in

general (Miller, 2012) and conduct assessments in particular (Harari et al., 2016). In recent years,

non-proctored online testing has become the main assessment mode in the academic field (Allen

& Seaman, 2014). The advantages of non-proctored testing can incur the cost of lack of

supervision, less standardized test-taking conditions, and less control over students’ behaviors.

The issues of dishonest behaviors in non-proctored assessments is then leading to unfair scores

and threats of the usefulness of online tests as a whole may arise (Steger et al., 2018). Proctoring

choices may be considered the responsibility of each individual institution to look into.
 38

At one institution, characteristics including students’ efforts and performances were

examined in regard to online assessments versus assessments administered in a classroom

proctored setting in the Spivey and McMillan (2014) study. The researchers measured students’

efforts by tracking the number of times students accessed the study resources in the university

system, and the study determined, students’ efforts and course performance were not influenced

by testing differences. Instead, success and students’ efforts in the classroom showed a strong

correlation with success on assessments. Mozes-Carmel and Gold (2009) determined the modality

used to administer the final exams in online courses did not have a significant difference. The

study indicated, sophisticated online software was a viable alternative to in-person proctored

exams as long as the academic rigor was taken into consideration and institutions minimized the

ability for students to access other online information (Mozes-Carmel & Gold, 2009).

According to Hollister and Berenson (2009), some instructors have been reluctant to teach

online because there is a concern about who is working on assessments. In a higher education

computer course study conducted by Hollister and Berenson (2009), the students’ scores in

proctored versus non-proctored exams were no different. The study concluded, the lack of

variation is due to the administration of questions in a simulated Microsoft Office environment

which is not conducive to cheating behavior. In another information systems course, Hollister and

Berenson (2009) analyzed the differences in online proctored versus non-proctored exams and

determined the non-proctored students had a significant amount of variation in performance,

although there was no significant difference in the mean overall performance. The researchers

determined the activity-based online exams graded by the computer were the determining factor;

hence, implying using these types of questions may help alleviate instructor concerns in regard to

online assessments.
 39

Administering the correct type of test is important to Brallier and Palm (2015), who

studied proctored and non-proctored tests in higher education online and face-to-face beginner

psychology classes. Although students scored 6% higher on the non-proctored online tests, the

course grades were ultimately not higher. Daffin and Jones (2018) conducted a study in

psychology classes with proctored and non-proctored tests and found the non-proctored tests were

10–20% higher, and students took twice as long to complete them. Test scores were similar for

the face-to-face and online courses for both proctored and non-proctored tests. So learning

outcomes in well-designed online courses are similar to those achieved in traditional courses

(Daffin & Jones, 2018). Due to the possible occurrence of misconduct during a non-proctored

test, one suggestion may be, the instructors should design the tests with the potential for

misconduct in mind and create questions which require far more than memorization (Brallier &

Palm, 2015).

Different subject areas in higher education may be questioned as to whether different

results would be achieved in a particular topic of study. Yates and Beaudrie (2009) studied

mathematics classes at the community college level, and determined the test results of proctored

and non-proctored online students had no significant difference with regard to grades. The study

included over 800 students, where approximately 400 students were administered proctored tests

and the rest were administered non-proctored tests online. A community college in New York

studied by Trenholm (2009) suggested, contrary to the proctored and non-proctored scores, there

may be a high correlation in online math test grades with institutional affluence, socioeconomic

status with computer accessibility, and student ability. Furthermore, each institution may be

looking at different concepts in regard to online courses.

 40

Integrity in Online Classes

The proctoring of tests may be a requirement for some who believe there are integrity

issues associated with taking a test online in the convenience of a student’s home. According to

Moore, Head, and Griffin (2017), some of the problems with online exams lie in the identification

of the test taker, the prevention of test theft, students’ use of unauthorized notes, cell phones,

and/or Bluetooth devices, and the determination of intentional computer crashes. If one question

becomes compromised, the integrity of the exam is affected. If the whole test becomes

compromised, then all the efforts put into the test as well as the test problems become damaged

(Moore et al., 2017). Instructors should note, Microsoft Word automatically creates a copy of

these tests on students’ computers when the documents are opened without any additional action

taken by the students. Ladyshewsky (2015) suggested the test design, the value of the test, student

age, and student maturity are the most important factors to consider when designing an

unsupervised online test. Ladyshewsky found cheating may not be as problematic as some

individuals fear. Rather, cheaters seem to cheat regardless of the testing modality.

Cifuentes and Janney (2016) explained, integrity concerns not only the students, but the

school as a whole. The promotion of integrity should be a foundational principle to which the use

of learning management system (LMS) preventive technologies should be applied; authentication

of students should be required, proctored exams should be in place, and students should be

required to submit individual work for each test. The Center for Innovation in Mathematics

(2011) concluded in a study, the proctoring of two exams (out of three or four) in addition to the

semester final exam provided enough motivation for students to ensure the learning outcomes

were met. Proctoring some assessments was the option used after data were analyzed because,

logistically, the proctoring of all tests might not be an option for many colleges. Since final exam
 41

grades were found to have a high correlation in overall average scores and knowledge of the

course, researchers determined the students needed to achieve 60% or higher on the final exam in

order to pass the class (The Center for Innovation in Mathematics, 2011).

Due to the elevated faculty concerns as well as more online classes and assessments being

offered, Varble and Haute (2014) conducted a study focused on reducing online cheating

opportunities, which compared online classes to traditional classes. In both cases, the last two

tests were proctored on campus along with the final exam. The data analysis results indicated

traditional students performed better when taking the proctored assessments and when there were

no aids available. A study conducted by Fask, Englander, and Wang (2015) in regard to proctored

and non-proctored testing in online statistics classes suggested, even though cheating in non-

proctored environments facilitates much higher levels of cheating, there are different inclinations

associated with student cheating. The study suggested the different disciplines, different levels of

courses (undergraduate, graduate, and professional), and different modes of interaction among

students and teachers can make a significant difference with integrity in online classes.

Alessio, et al., (2017) addressed the challenge of online testing and the compromising of

tests due to undetected cheating which can produce inappropriate higher grades. Tests of 147

students in multiple sections of online courses were compared with half of the students who

received no proctoring, while the other half utilized online proctoring software. On average,

students scored seven points lower and utilized a significantly smaller amount of time to take

online tests when the tests were proctored. Another challenge and concern, according to Alessio

et al. (2017), is the attrition rate, since 7% of the students with non-proctored tests dropped

classes, while 19% of the students with proctored tests dropped classes. Samavati, Stumph, and

Dilts (2012) determined the students who scored higher on tests had a negative correlation with
 42

time and score, which suggests, in online environments, the allocation of extra time does not

necessarily help improve a student’s grade. Hence, to improve online assessment outcomes,

instructors are recommended to consider reducing the allotted time to about 75% of what is

allowed for a comparable in-class proctored test (Samavati et al., 2012).

Unethical behavior and dishonesty are widespread in the public and private sectors, which

cause not only integrity issues, but large financial losses (Ayal, Gino, Barkan, & Ariely, 2015). In

education, reminding students to refrain from cheating can be an effective practice. The reminders

may increases salience while decreasing the ability to justify. Self-engagement, which increases

students’ motivation to maintain a positive self-perception, can bridge the gap between moral

values and behavior. With all the above in place, Fask, Englander, and Wang (2014) stated, if a

difference in online versus traditional test scores are observed, establishing whether the variance

is due to cheating or the modality of the test (in class versus online) is still important.

Online cheating internet companies who support academic dishonesty by writing papers,

completing homework assignments, and enrolling on behalf of students pose a significant

challenge (Moten, Fitterer, Brazier, Leonard, & Brown, 2013). Due to these challenges, the study

recommended online assessments and courses to be designed, knowing students use textbooks

and notes in order to ensure more comprehensive courses. The assistance from the internet in

regard to cheating has been a game changer compared to years ago, according to Watson and

Sottile (2015). Programs such as Turnitin.com have assisted with plagiarism; due to proctoring

programs, instructors may place more weight on projects and papers versus tests taken for online

classes (Watson & Sottile, 2015).

Other suggestions given by Moten et al. (2013) included the creation of multiple versions

of assessments, the randomization of questions, or a requirement to have online students sign and
 43

return an academic dishonesty statement. Cheating among freshmen and graduate students was

found to be the highest though the study, according to the Moten et al. (2013), and need to be

looked into. Some people may claim, the freshmen dropped out, but dropping out is not an excuse

to be considered viable in the case of graduate students. More research should be conducted on

online assessments (Watson & Sottile, 2015).

Due to the growing popularity of online courses, tech-savvy students can still find ways to

cheat, which allows the student to pass online classes without much effort while remaining

difficult to detect. The phenomenon was described by one student surveyed in the Miller and

Young-Jones (2012) study, who received an “A” grade in his online course because four friends

shared a Google Document online, which all students could read and add to simultaneously. Due

to the surveys administered in the study, companies were suggested to work on developing new

and improved anti-cheating software, and researchers and schools alike need to join forces and

share work. Unless sharing occurs, schools and instructors are going to continue playing the

catch-up game (Miller & Young-Jones, 2012). Some technologies, according to Karim,

Kaminsky, and Behrend (2014), have already been examined to determine effectiveness regarding

cheating, although have not determined whether there even exist unintended effects on students’

reactions, performances, and/or selection procedures.

Karim et al. (2014) conducted a study in which students were assigned randomly to a

webcam-proctored or honor system condition for two online tests. Regarding the 295 students

who participated, the researchers determined remote proctoring may have reduced cheating rates,

although unintended student reactions may have increased, which might differentially affect a

test-taking experience. Institutions who test both anxiety increases and coping skills are needed.
 44

The data invites the opportunity for instructors to incorporate material into online courses to help

students become calmer while taking proctored online exams (Kolski & Weible, 2018).

Companies who have started to offer proctoring services for online classes are faced with

an uphill battle in persuading skeptics, as cameras remotely peer into a student’s home, seize

control of the computer, and stare at the student for the duration of the test to determine whether

there are any signs of impropriety (Kolowich, 2013). Each company has an individual approach,

such as using webcams, sharing computer monitors, recording a student, and installing software

which makes using other browsers or chat platforms impossible. Stack (2015) conducted a study

with 287 online criminology students who took assessments either proctored at the college or

proctored at home using the Respondus Lockdown Browser. The study determined, there was no

significant difference between student exam scores in either sections, which suggests, online

proctoring systems may begin to level the playing field. Kolowich (2013) claimed the real

question revolves around how much proctoring is enough in higher education where institutions

certify academic achievement. In general, the study concluded, online proctoring needs to be at

least as effective as what one would witness in a large lecture classroom (Kolowich, 2013).

Because higher education learning management systems, such as Blackboard and Canvas,

are becoming the norm, computerized testing has made testing not only easier for instructors, but

more convenient for students (Tao & Li, 2012). For instance, the cost of delivery is cheaper, there

is improved efficiency of administration, the time-consuming scoring is completed by the

computer, and more time can be saved for interactions with the instructors. When students know

they are going to be proctored during an online assessment, the assessment is taken more

seriously and frequently more studying is involved. Stowell (2015) studied a biological

psychology class for three semesters to delve into the effects of open-book non-proctored online
 45

testing for students and instructors. As expected, students scored higher, but the effect was only

temporary and did not last throughout the semester.

Trends in increasing online courses at institutions suggest the need for further research on

cheating and academic integrity in comparison with those of face-to-face classes. The Miller and

Young-Jones (2012) study surveyed 639 students in each type of course. Although students felt

cheating was easier in online courses, those who only took online classes were less likely to cheat

than students who only took face-to-face courses. For students taking both online and face-to-face

courses, cheating did occur more frequently in online courses. Older students tended to cheat less

frequently, as well, but in regard to female versus male students, no significant differences were

identified. In the Swartz and Cole (2013) study in which undergraduate and graduate students

were surveyed, the majority of students believed academic integrity possessed the same degree of

issues in online environments as students did in face-to-face courses when taking an exam,

writing a paper, or completing a project. Additionally, the students noted, if someone wants to

cheat, students are going to find a way to cheat no matter the type of environment (Swartz &

Cole, 2013).

Chapter Summary

Chapter 2 literature review examined possible effects on assessments, success, and

retention in online mathematics classes through the literature search strategies and frameworks.

Topics discussed in the research were online classes in higher education, success of students in

online classes, retention, the number of students passing proctored versus non-proctored tests, and

integrity in online assessments. The theoretical framework for the research study is based on the

theory of planned behavior, which aims to predict and understand the relationship between human

behavior and motivation (Boslaugh, 2013). Demographic variables, included in the theory, are
 46

believed to contribute a significant influence on outcomes related to cheating behaviors (Madara

et al., 2016). Planned behavior theory presents a reason as to why students may violate ethical

codes under these circumstances people would otherwise support (McQuillan & Zito, 2011). A

gap in the literature lies specifically in the data from online mathematics classes, as other subject

areas do not apply for proctored and non-proctored testing. The proposed study is necessary

should colleges hope to increase online success and retention.

A common theme emerged, which suggested integrity in online assessments should be

researched within individual institutions. Integrity is essential to the flexibility, consistency,

reputation, and overall survival of higher education and online education so the reduction may

lead to shocking consequences for the future of higher education (Farnesea, Tramontanob, Fidaa,

& Pacielloc, 2011). Many students perceive a college degree as being a pass needed to enter into

the attractive middle- or upper-class lifestyle, and the pressures to succeed may lead to academic

dishonesty when the achievement of the goal is put at risk (Farnesea et al., 2011). The presence of

academic dishonesty is indisputably present in all cultures and should be minimized as much as

possible in the education field as a whole. The observable differences lie in the scope of

dishonesty, the attitudes toward dishonesty, and the penalties which result from dishonesty

(Blachnio & Weremko, 2011).

Murdock, Hale, and Weber (2001) reported an increase in cheating over the last few

decades, as well as a diminished trend in students’ perceived seriousness of dishonest behavior.

The Mathematics Department at the Mid-Atlantic community college is concerned about the

integrity of students who are taking all assessments online in the mathematics department.

Integrity can equally affect the students and the institutions. The study was aimed to determine

whether final exam scores, success (passing with a 70% or above for the semester), and retention
 47

(signing up for the subsequent math course) are affected when assessments are proctored versus

not proctored during the semester. Chapter 3 addresses the research methods used in the study.
 48

Chapter 3: Methodology

The purpose of the quantitative research design was to examine the relationship of final

exam scores (pass/fail), percent of student success (passing the semester with a 70% or higher),

and percent of student retention of online mathematics students at the Mid-Atlantic community

college (MACC). Final exam scores, percent of success, and percent of retention were compared

between proctored versus non-proctored tests during the semester to determine if there is a

relationship between the data. The data was also broken up according to age and gender. The

study was necessary in order to determine if the college can increase online success and decrease

dropout rates through the type of testing given to students during the semester.

Percent of student success in online math courses with proctored testing were compared

with the percent of student success in online math courses with non-proctored testing. A series of

logistic generalized linear mixed models (LGMM) compared the percent of student retention in

online math courses with tests proctored during the semester versus those who had non-proctored

testing. Final exam scores in online math courses with proctored testing were also compared with

final exam scores in online math courses with non-proctored testing. Data were collected from the

Program Research and Evaluation (PRE) Department and organized for accurate research.

Statistical analysis started with descriptive statistics to describe percentages of variables.

Continuous variables were reported using mean and standard deviation.

A series of logistic generalized linear mixed models examined the relationship

demographics and variables. The tests were conducted to examine the relationships between

proctor status and all the dependent variables. Mixed models were used in order to account for the

random effect of courses and subjects who attended different terms multiple times. Proctor was
 49

used as a predictor for all outcomes controlling for gender, race, and age. Data were managed and

analyzed using IBM SPSS version 25.

The tests were used to determine whether there was a significant difference between what

is expected in online courses and the observed frequencies among the variables. The test were

then used to determine if the null hypothesis should be rejected (University of Pennsylvania

School of Arts & Sciences, 2008). In the study, the variables are success, retention, and final

exam scores. Materials collected are kept in a password-protected folder on a personal computer

for three years after the completion of the study.

Research Questions and Hypotheses

The following hypotheses were developed based on the research questions for the study.

Hypotheses were written based on the quantitative research design using a series of logistic

generalized linear mixed models for the statistics for analysis. The following research questions

and hypotheses guided the study:

Research Question 1: What is the relationship between student final exam scores,

(pass/fail) for students with proctored versus non-proctored online math tests?

Research Question 2: What is the relationship between the percent of student success for

proctored versus non-proctored online math tests?

Research Question 3: What is the relationship between the percent of student retention for

proctored versus non-proctored online math tests?

H10: There is no relationship between final exam scores (pass/fail) of students with

proctored versus non-proctored online tests.

H1A: There is a relationship between final exam scores (pass/fail) of students with

proctored versus non-proctored online tests.

 50

H20: There is no relationship between the percent of student success with proctored versus

non-proctored tests.

H2A: There is a relationship between the percent of student success with proctored versus

non-proctored tests.

H30: There is no relationship between the percent of student retention with proctored

versus non-proctored tests.

H3A: There is a relationship between the percent of student retention with proctored versus

non-proctored tests.

Research Design and Rationale

The research study used final exam scores (pass/fail), success, and retention of students in

mathematics classes. The research design included a series of logistic generalized linear mixed

models to determine whether there was a significant difference between the expected occurrences

and the observed occurrences. A series of logistic generalized linear mixed models (GLMM) were

conducted to examine the relationships between proctor status and all the dependent variables.

Mixed models were used in order to account for the random effect of courses and subjects who

attended different terms multiple times. Proctor was used as a predictor for all outcomes

controlling for gender, race, and age. Hypotheses for the study determined if there were

relationships between final exam scores, the percent of success, and the percent of retention with

students who have assessments proctored and non-proctored throughout the semester.

Quantitative research and hypotheses questions rely specifically on directional language, and

focus on the variables under investigation, the relationship to each other, and affect different

groups (Leavy, 2017).

 51

Quantitative research design is the appropriate design for the study because the design

focuses on the following variables under investigation, including final exam scores, success, and

retention, which are under investigation with proctored and non-proctored testing. Logistic

generalized linear mixed models (GLMM) were used to determine if proctoring student tests

during the semester would be better practice. Given the variables are independent, some call the

test a "goodness of fit" statistic, because the test measures how well the observed distribution of

data fits with the distribution expected (Deshpande, 2011). The results helped explain how future

test in online mathematics courses should be given. Quantitative research design is used to

advance knowledge in education by identifying relationships between two variables (Creswell,

2017).

Research Procedures

The discussion of research procedures includes population and sample selection,

instrumentation, data collection, and data preparation. Population, which entails all online

mathematics students at MACC, is a collection of elements and data in a specific region at a

certain point in time, which are considered the subject in the study. A student, for instance, in an

online math course is a unit of the population. The sample design is specifically online math

classes in the past three years.

These classes were chosen because each one was an institutionalized online mathematics

class MACC offered. The classes included uniformed online assessments and finals, where the

only difference was the randomly assigned numbers (of the same difficulty level) which were

inserted into the tests. The instrumentation section explains how the PRE Department collected all

of the data for the study and describes how the data were used to benefit the research. Research
 52

indicates readily available student data within institutions can be an effective predictor for student

retention and success (Gomez, 2013).

Population and Sample Selection

The Mid-Atlantic community college serves over 62,000 students with about 40% of the

students being white, 40% African American, 6% Hispanic or Latino, 8% Asian, 4%

multicultural, and 1% unknown. MACC students who take at least one online course make up

about 31% of the whole population. There are approximately 43% white, 39% African American,

6% Hispanic or Latino, 7% Asian, 4% multicultural, and 1% unknown students in online classes.

In addition, there are over 35% of students who are Pell Grants recipients. The population of the

study was approximately 1,900 online mathematics students which MACC serves yearly. The

data in the study were a comprise of data from students in online institutionalized mathematics

courses at MACC in the past three years. These classes are created by a team of faculty members,

approved by the administration, and Quality Matters (2018) approved. The course is then given to

each instructor to teach for consistency.

Secondary data from students in each of these online mathematics classes was collected as

numeric and non-numeric data which can answer new research questions on data already

collected (MacInnes, 2017). Any student who dropped the class (no longer registered) early

enough to be taken off the roster in the first week, was not be included. Any students who receive

a withdraw (W) grade, were included and recorded as failing for the semester with a 0% final

exam grade. Permission was given from the administration (see Appendix A), and data were

collected from the PRE Department. The data came from 10 different instructors who teach online

mathematics courses. There were about about 300 students each semester, for three years, which

totals about 1900 students. Within these classes, there are sections which have no proctored tests
 53

during the semester, and only the final exam is proctored, and other sections which do have

proctored tests along with the final exam.

A sample is a subset of the population and is important to ensure the sample size is

appropriate to make inferences about the results of the study (Hoy & Adams, 2016). The

predictors in the model are proctoring status (proctored tests for online classes, non-proctored test

for online classes, and traditional proctored test for face-to-face classes) combined with gender,

age, and five different math classes in which students could have taken proctored tests in. When

analyzing data, the representing sample should have a well-defined population. The intent is to be

able to make a generalization from the sample to the population (Creswell, 2017).

In summary, after getting permission from administration, a sample of the population of

online mathematics classes was used in the study. Data were collected from the classes in the past

three years, on final exam scores, success, and retention to see if there was a relationship with

students having proctored versus non-proctored tests during the semester. With the data, a series

of logistic generalized linear mixed models tests were conducted with an effective sample size for

the research, which is necessary to obtain a desired level of statistical power (Anderson, Kelley,

& Maxwekk, 2017). The percent of students who achieved passing exam scores, end of semester

success, and retention were compared in charts. After the research was complete, what should

happen could be extrapolated from the entire population in reference to online mathematics

courses and proctored tests.

Instrumentation

Final exam scores (pass/fail), success, and retention were categorized into chart form

according to students who had proctored versus non-proctored testing during the semester. Data

were collected from the PRE Department on students in face-to-face math courses to compare
 54

percentage of success as well. Then, by using a series of logistic generalized linear mixed models

tests on the collected data, the determination was made if proctored versus non-proctored testing

during the semester, in online mathematics courses, makes a difference in final exam scores,

success, and retention, in order to answer the research questions.

Final exam scores, success, and retention were dichotomous data written as Pass/Fail and

Retained/Not Retained as well as percent of passing and retained. To establish reliability, there

was internal consistency. The online mathematics courses each instructor uses were designed by a

team, approved by the administration, and Quality Matters Certified. Quality Matters Certification

(2018) is obtained when the company has reviewed the course, and certified the course as well-

designed, engaging for learners, and has met standards for the disabled. These courses have all the

homework, tests and final exams in the same format, with the same questions, and randomly

assigned numbers of the same difficulty level inserted into each question. The Mathematics

Department and administration have determined, students who are successful on the final exam in

online mathematics courses are prepared for the subsequent mathematics course, which is an

educational goal for all.

Data Collection

Data collection according to Zozus (2017) include identifying data to be collected,

defining the data elements, measuring values or acquiring secondary data, processing data in

electronic forms to prepare for analyzing. In order to prepare for correlational quantitative

research, data were collected from MACC instructors and the schools Planning, Research and

Evaluation (PRE) Department after approval has been given. Access was given to the data for the

study and the data were emailed from the PRE department to conduct the study. Only online

mathematics classes in the past three years were a part of the data collection. The data were
 55

collected from 16 different instructors for three years, totaling approximately 1900 students.

There were classes with only a proctored final exam, and others with proctored tests as well as the

final.

The test scores, passing percentages, and retention were given to the instructor in the form

of a student ID. These were then converted to a randomly assigned and anonymous

numbering/coding system. These identification numbers ensured confidentiality to all involved.

The data is housed on a computer in a password-protected folder. Researchers need to remember

validity relies on accurate and truthful data collected without fear of disclosure of sensitive

information (Adinoff, Conley, Taylor, & Chezem, 2013).

Mid-Atlantic community college instructors use an open educational online assessment

system called MyOpenMath (2018). All final exams are taken in the system and stored here

online. Once the college approved the process, an administrator allowed the access to the final

exam scores in the system. The success percentages of online students and face-to-face students

were collected through the PRE department at the college. PRE was told to ignore any one credit

online courses, which occur when students only need to finish one portion of a course.

The PRE department provided a spreadsheet of data on each of the original students in the

study who signed up for a subsequent mathematics course. The original students were tracked by

PRE to see if the students signed up for another course after being successful in the previous

course. To ensure participant confidentiality, students were coded with an identification number

to ensure confidentiality. Even when data is collected using standardized procedures and tools, the

data still needs to be checked for possible inaccurate or missing data to ensure the findings are

clear and conclusions can be validated, and verified (Peersman, 2014). Any problems which

arouse in the process of data collection should be noted along with how the problems were
 56

overcome, in order to keep a transparent nature for anyone who would want to review the process

(Best, 2014).

The student’s mean and median final exam grade were determined with given standard

deviations, and the percent of student success at a 70% semester average or higher, given the type

of online class signed up for. The series of logistic generalized linear mixed models determined if

the variables had a relationship. These tests allowed inferences about means to be made by

analyzing variances to find out if there was a relationship between proctored testing, non-

proctored testing and student success.

Data Preparation

During the data preparation period, there was the collection of data, consolidation of all

data, and putting the information into tables to be analyzed (Zozus, 2017). The data entailed

coding, entering data into the computer for analysis, checking and verifying all elements are

accurate (Cohen, Lawrence, & Morrison, 2017). First, final exam scores were recorded in chart

form from highest to lowest and organized by students who had proctored or non-proctored non-

final assessments during the semester. Second, to determine success, the students needed to have

a 70% or higher as a final average for the course. At MACC, a 70-79% is a C, an 80-89% is a B,

and a 90-100% is an A. A chart was designed to list students who were successful versus

unsuccessful in face-to-face classes, which have all proctored testing, online classes with

proctored tests, and online classes without any proctored assessments during the semester. There

is a concern for success in online higher education classes because students are often unable to

socially interact with peers and instructors, as face-to-face peers do, which ultimately helps in the

learning and engaging process (Struble, 2014). The third step, to determine retention percentages
 57

among given groups, is a chart to classify students who did or did not sign up for the subsequent

mathematics course.

Data Analysis

Data analysis is about asking questions, developing explanations, and testing hypotheses

based on logic. Data analysis is multidisciplinary, and combines artificial intelligence, statistics,

math, and business (Cuesta & Kumar, 2016). Data analysis has been discussed below with

reference to each research question of the study. Each separate discussion of data analysis

includes the rules for hypothesis testing. A series of logistic generalized linear mixed models was

performed to determine if there was a relationship between variables in the research questions.

The significance of the results was then calculated. If the calculated value is larger than the

critical value, the null hypothesis was rejected, which suggests a significant relationship

(Statistics Solutions, 2013). In order to analyze the data correctly, there needs to be screening

procedures. Initially, data had to be checked for mistakes, and abnormalities, since data can easily

be entered incorrectly. Outliers need to be taken into consideration as well, because outliers can

have a high impact on statistical analysis. Being very familiar with the data collected and

understanding the distribution of each variable through the examination of descriptive statistics, in

order to easily detect if anything appears to be out of the ordinary, the research may be highly

benefit (DeSimone, Harms, & DeSimone, 2015).

Data Analysis, Research Question 1

The first research question of the study is as follows: Is there a statistically significant

effect of proctoring status on the scores of students’ final exams in online mathematics courses?

The independent variable, according to the University of Connecticut (2018), is a factor which is

selected by the experimenter to determine the relationship in a study. In the study, the variable is
 58

reliant on a student in a test proctored or non-proctored online class, and the dependent variable is

a student’s score on the final test. RQ1 data analysis contained categorical variables to track each

student’s gender, age, and the exact class the students took. The classes were also be compared

with percentages and letter grades.

The dependent variable of final score, the independent variable of proctoring status, and

the covariates of gender, age, and math class determined how proctoring effects final exam

scores. There are two different proctoring states, the presence of proctoring during semester tests

or the absence of proctoring during semester tests, on the final score (final). The tests determined

if there is a difference in final exam scores in proctored versus non-proctored tests for students.

Data Analysis, Research Question 2

The second research question of the study is as follows: Is there a statistically significant

effect on the proctoring environment on students’ success? The purpose of data analysis for

Research Question 2 (RQ2) was to determine whether the proctoring environment might be

associated with better overall class performance for community college mathematics students. In

RQ2, the independent variable is proctoring, and the dependent variable were reliant on a student

obtaining a C or better (>70%) in the class. RQ2 data analysis contained categorical variables to

track a student’s gender, age, and the exact class the student took. The test determined if there is a

difference in success with proctored versus non-proctored tests for students.

Data Analysis, Research Question 3

The third research question of the study is as follows: Is there a statistically significant

effect of proctoring status on the retention of students in online math courses? The purpose of

data analysis for Research Question 3 was to determine whether proctoring might be associated

with better retention for community college mathematics students. In Research Question 3 (RQ3),
 59

the independent variable was reliant on a student being proctored in the online class, and the

dependent variable was reliant on a student being enrolled at MACC in another math class, the

semesters after taking proctored tests during the semester. Data analysis for RQ3 was restricted to

those students who were eligible for retention in the semesters after taking an online math course

with proctored testing; thus, students scheduled to graduate in the first semester after a proctored

test were excluded from the data analysis for RQ3. Finally, RQ3 data analysis contained

categorical variables to track a student’s gender, age, and the exact class the student took, and

percentages of retained students.

Reliability and Validity

When selecting and evaluating an assessment tool, one needs to consider reliability and

validity. Reliability refers to consistency (Jackson, 2015). When an assessment is reliable, a

consistency of scores is demonstrated. A consistency of scores occurs when scores obtained are

the same after the data is reexamined and approximately the same score is reached. Validity

determines if the research measures what was intended to measure and how true the results are

(Jackson, 2015). In statistical tests, both reliability and validity are dependent on the normality of

the distribution of the dependent variable (Jackson, 2015).

Potential threats to internal validity include items such as maturation, and evaluation

anxiety (Houser, 2015). Threats such as these can often be reduced or eliminated by choosing

certain research designs. Possible threats to quantitative research and internal validity can

compromise assurance in stating a relationship exists or does not exist with the dependent and

independent variable (Houser, 2015). A threat to internal validity was ultimately be an influence

on the results as well and generalized cannot be made (Cotrell & McKenzie, 2011). In the study,

dealing with students at the community college level, maturation could affect the study. Students
 60

can change over the course with physical or mental maturation changing scores Evaluation

anxiety affects students’ scores when students know they are being evaluated on a test and feel

uneasy.

External validity threats occur when incorrect inferences from data, settings, and past or

future situations are made. A threat could be population validity, which is generalizing the results

of a study of a specific group to another group or a larger group (Creswell, 2017). Using the

results of the study is important to all present and future online mathematics courses at MACC.

Ecological validity can become a threat if the study is not sufficiently described and conducted,

because of the difficultly in determining if the results apply to any other settings or groups, like

all online math classes in the study. Another threat could be interaction of history and treatment,

which occurs since studies are time bound, and the past and future cannot always be generalized.

The threat can be resolved by replicating a study at a later date. In the study, the final exam

scores, success, and retention of online mathematics courses could be studied again to verify the

results obtained (Creswell, 2017). Mathematics can be difficult for students, so looking at the

subject specifically in face-to-face classes is important, where an instructor is going through

examples together with the class, versus online where there are videos for only some examples, is

important. Another way to think about validity and to ensure threats are limited, is to understand

the validity of tests is a combined judgment of empirical evidence and theoretical rationales

which support the adequacy and appropriateness of implications and actions based on test scores

(Khemakhem, 2016).

Ethical Procedures

There are legal and ethical obligation to keep all the human subjects involved, protected

and confidential. As Bain (2017) acknowledged, ethical approval should be a moral reflex for all
 61

researchers, and ethical training should be provided when needed. The data collected is housed on

the researcher’s computer in a password-protected folder. The students are coded by giving

identification numbers to ensure participant confidentiality is protected. To increase the

credibility of the data, and validity, all online mathematics courses which fit the criteria were

used. Multiple full-time online instructors, adjunct online instructors, and administrators were

contacted to determine if proctored or non-proctored tests during the semester were given in each

class. Researchers need to keep perspective of others for ethical practice as instructors of a

university. Even though instructors/researchers may be given access to data like grades and test

scores for a specific reason, permission is still needed to use part or all of the data for any research

purpose (McGinn, 2018).

Chapter Summary

The purpose of the quantitative design was to analyze scores, percent passing, and percent

of retention in online mathematics students at the Mid-Atlantic community college (MACC).

Quantitative research focuses on the variables under investigation and the relationship each has to

the other (Leavy, 2017). Data from the study determined if proctored assessments affect MACC

students in online math courses when determining success, with a passing rate of a C or above.

The results of the study were discussed in the Mathematics Department and decisions can be

made about online mathematics classes, when determining regulations with assessments for the

purpose of improving success and retention. Retention is upheld at the college when students who

have successfully completed a mathematics course sign up for the subsequent course needed.

Success in online education, as Young, Birtolo and McElman (2009) explained, is understanding

who the customer is and understanding what the customer needs.

 62

Discussed in chapter 3 are the research design, rationale, and procedures where the

population and size are defined along with the sampling strategy. The chapter included who the

participants were, and how consent was obtained. In the data collection techniques section, the

data were collected from final exam scores, the percent of success, and the percent of retention.

The appropriateness, reliability, and validity were discussed. The data section included the

procedures for participation, data collection, gaining access to the data set, and permission. The

data collection part of the chapter is described, how the dependent and independent variables are

collected, and how the data is stored safely and confidentially even after the study. Even though

there has been mention of education plans to make data collection inventory publicly available

through a searchable web database, for some fields, the data is still all confidential (Scott, 2013).

In Chapter 4, the data analysis section, which can study growth over time using the same

variables provided explanations of data (Nese, Lai, & Anderson, 2013), and statistical tests used

to test the hypotheses, rationale, and interpreted results. Describing reliability and validity has

been done by identifying internal and external threats. The ethical procedures are then described

how the human participants are protected ethically throughout the study and after. Ethical

procedures are all upheld by making the data anonymous and confidential.
 63

Chapter 4: Research Findings and Data Analysis Results

The purpose of the quantitative research was to examine whether proctoring exams in

online math courses is associated with final exam scores, the final course grade for success, and

retention in the areas surrounding Baltimore, Maryland. To achieve the purpose of the study, three

research questions were addressed:

Research Question 1: What is the relationship between student final exam scores,

(pass/fail) for students with proctored versus non-proctored online math tests?

Research Question 2: What is the relationship between the percent of student success for

proctored versus non-proctored online math tests?

Research Question 3: What is the relationship between the percent of student retention for

proctored versus non-proctored online math tests?

H10: There is no relationship between final exam scores (pass/fail) of students with

proctored versus non-proctored online tests.

H1A: There is a relationship between final exam scores (pass/fail) of students with

proctored versus non-proctored online tests.

H20: There is no relationship between the percent of student success with proctored versus

non-proctored tests.

H2A: There is a relationship between the percent of student success with proctored versus

non-proctored tests.

H30: There is no relationship between the percent of student retention with proctored

versus non-proctored tests.

H3A: There is a relationship between the percent of student retention with proctored versus

non-proctored tests.
 64

Data Collection

A total of 1909 cases (students) across five different levels of online mathematics courses

at a Mid-Atlantic Community College (MACC) from 2016 to 2018 were included in the study.

The largest population possible was collected of complete data (final exam scores and final course

grades). In 2016 the community college implemented institutionalized online mathematics

courses so students would have more of the same experience, with a course which met all of the

school’s standards and expectations. Each instructor teaches these courses which included

uniformed online assessments and finals, where the only difference was the randomly assigned

numbers (of the same difficulty level) which were inserted into the individual test items.

The Mid-Atlantic community college serves over 62,000 students with about 40% of the

students being white, 40% African American, 6% Hispanic/Latino, 8% Asian, 4% multicultural,

and 1% unknown. MACC students who take at least one online course make up about 31% of the

whole population. There are approximately 43% white, 39% African American, 6% Hispanic or

Latino, 7% Asian, 4% multicultural, and 1% unknown students in online classes. In addition,

there are over 35% of students who are Pell Grants recipients.

Secondary data from students in each of these online mathematics classes was collected as

numeric and non-numeric data which were used to answer research questions regarding the data

already collected (MacInnes, 2017). Any student who dropped the class (no longer registered)

early enough was taken off the roster in the first week and were not included. Any students who

received a withdraw (W) grade was included and recorded as failing for the semester with a 0%

final exam grade. When dealing comes to financial aid, a W is considered a fail (F). Permission

was given by the administration (see Appendix A), and data were collected from the Planning

Research and Evaluation (PRE) Department. The data came from 16 different instructors who
 65

taught the online mathematics courses. For the previous three years, the total students with both

pre- and post- test scores is about 1900. Within these classes, there are sections which have no

proctored tests during the semester, and only the final exam is proctored, and other sections which

do have proctored tests along with the final exam. The online instructors were surveyed to

determine if the instructor had proctored assessments or non-proctored in each class.

At the time of the study, students attending the MACC were allowed to register for online

courses. The study was conducted with five different levels of mathematics courses and over

three years. Students who were in classes with proctored assessments during the semester were

compared to those who did not have proctored tests during the regular semester. In order to do

this, final exam scores (pass/fail), success rates (pass/fail) and retention rates were compared. The

three years includes winter, spring, summer and fall online courses. Therefore, these approximate

1900 students are representative and proportional to the larger population. All online mathematics

courses and future online mathematics classes would be included at the community college for

external validity.

Data Analysis and Results

Prior to analyzing the data, all data were cleaned: duplicates and impossible (extreme –

disparate) values were examined first. The examination process found 282 cases were duplicated,

meaning some subjects were included in the data multiple times at different terms for different or

same courses. For instance, if a student failed one of these courses, the student was counted as

being not successful in the course final grade. If a student signed up for the course again (possibly

with another instructor), and passed the course, students were counted as being successful in the

class for the semester. Another situation could be if a student passed one of these online math

courses, and then signed up for another subsequent online class in the study, and passed, students
 66

were counted in both situations. There were 2.2% of missing data, but the data were missing

completely at random (MCAR), which means there is no relationship between whether a data

point is missing and any values in the data set (Sweet & Grace-Martin, 2012). Pairwise deletion

was then used, missing and duplicates were removed prior to the statistical analysis. Statistical

analysis started with descriptive statistics to describe frequencies and percentages of all

categorical variables. Continuous variables were reported using mean and standard deviation.

Next, cross tabulations using chi-square tests and independent t-tests were conducted to examine

the relationships between demographics and independent/dependent variables in order to

determine if any covariates needed to be included in the analysis. In the primary analyses, a series

of logistic generalized linear mixed models (GLMM) were conducted to examine the

relationships between proctor status and all the dependent variables. Mixed models were used in

order to account for the random effect of courses and subjects who attended different terms

multiple times. Proctor was used as a predictor for all outcomes controlling for gender, race, and

age. Data were managed and analyzed using IBM SPSS version 25. Alpha levels were set at α =

.05 and odds ratios (OR) were used as effect sizes.

Descriptions of the Sample:

A total of 1909 cases across five courses from 2016 to 2018 were included in the study.

Frequencies and percentages for categorical variables are displayed in Table 2. Gender, course ID,

Age, and race were all described.

 67

Table 2

Descriptive Statistics for Demographic Variables

Demographic variable n %

Gender
Female 1367 71.6
Male 539 28.2

Race
African-American 776 40.6
American Indian/Alaska Native 6 .3
Asian 82 4.3
Hispanic/Latino 70 3.7
Multiple Races 93 4.9
Native Hawaiian or Other PI 3 .2
White 857 44.9

Course ID
MATH081 (Pre-Algebra) 179 9.4
MATH082 (Introductory Algebra) 541 28.3
MATH083 (Intermediate Algebra) 517 27.1
MATH163 (College Algebra) 378 19.8
MATH165 (Pre-Calculus) 211 11.1

M SD
Age 29.0 8.2
_________________________________________________________________
Note. Frequencies not summing to 1909 reflect missing data.
A majority of cases were females (71.6%) with an average age of 29 years old. The largest

percent of the students were White (44.9%), followed by African-American (40.6%).

Only 1227 cases provided proctor information, of which 68.2% (43.8% of total) were

proctored. Most cases failed the final exam (68.9%), with mean final grade of 40.2 (SD = 34.7).

However, nearly half of the sample succeeded for the course (48.2%) and nearly half of the

sample retained in the math (49.6%). More details describing independent and dependent

variables are in Table 3.

 68

Table 3

Descriptive Statistics for Independent and Dependent Variables

_______________________________________________________________
Independent/Dependent variable n %
Proctor
Not proctored 390 20.4
Proctored 837 43.8

Final exam
Fail 1315 68.9
Pass 507 26.6

Course grade
Not success 905 47.4
Success 921 48.2

Retention
Not retained 879 46.0
Retained 947 49.6

M SD
Final exam grade 40.2 34.7
_________________________________________________________________
Note. Frequencies not summing to 1909 reflect missing data.

Proctor Comparisons for Final Exam, Course Grade, and Retention:

Several logistic generalized linear mixed models (GLMM) were used to examine the

relationships between proctor and outcome variables. The models account for the random effect

of course and subjects who attended multiple terms. Logistic regression is used because the

outcome is a binomial variable is determined (pass/not pass). From the preliminary analyses via

cross-tabulation using chi-square tests, the data demonstrated African-American students were

more likely to fail the final exam and had the lowest success rate. Males were more likely to

succeed than females, and older participants were more likely to succeed and be retained.

Therefore, the GLMM regressions controlled for age, gender, and race/ethnicity.
 69

Final exam scores of 70% or above were considered passing. Therefore, the final exam

grade was dichotomized into pass and fail. Results, shown in Table 3, and revealed students who

took the exams in a proctored class during the semester were more likely to fail the final exam,

OR = .669 (CI = [.486, .920]), p = .013 while controlling for age, gender, and race as well as

random effects for the course. The results indicate proctoring semester assessments was 1.49

(1/.669) times more likely to fail the final exam as compared to not proctoring semester

assessments. Whites and other races had higher success rates on the final exam than African-

Americans, OR = 1.775 and 1.930, ps < .005. As seen in Table 2, with race distribution, most

participants were White and African American. Although the other races are important, because

the individual sample percentage is below 10%, students could not be included in the regression

model, and were combined as “other.” Older participants were more likely to pass the final exam,

OR = 1.026 (CI = [1.010, 1.044]), p = .002.

Table 4

Logistic Regression Statistics with Proctor and Demographics Predicting Final Exam
______________________________________________________________________________
95% CI for OR
Predictor B OR p LL UL

Proctor a -.402 .669 .013 .486 .920

Race b
Other races .657 1.930 .004 1.234 3.017
White .574 1.775 <.001 1.305 2.413
Male c .136 1.146 .388 .841 1.561
Age .026 1.026 .002 1.010 1.044
______________________________________________________________________________
Note. a Compared to Not proctored. b Compared to African American. c Compared to Female.

When determining course grades for success, a participant who obtained a C or above

letter grade was considered passing the course. Therefore, course grade was dichotomized into
 70

success and not success. As seen in Table 5, students who were proctored were less likely to

succeed, OR = .652 (CI = [.494, .856]), p = .002 while controlling for age, gender, and race as

well as random effects for course. The result suggests students who attended the non-proctored

assessments during the semester were 1.53(1/.652) times more likely to succeed in the course than

the student who attended the proctored assessments. In addition, White and other races were more

likely to succeed than African-American, OR = 2.440 and 1.600, ps < .05. Older participants were

more likely to be successful for the course, OR = 1.022 (CI = [1.007, 1.038]), p = .004. Males

were more likely to succeed in a course than females, OR = 1.321 (CI = [1.006, 1.736]), p = .045.

Table 5

Logistic Regression Statistics with Proctor and Demographics Predicting Course Grade for
Success
_______________________________________________________________________
95% CI for OR
Predictor B OR p LL UL

Proctor a -.428 .652 .002 .494 .859

Race b
Other races .470 1.600 .017 1.088 2.354
White .892 2.440 .000 1.875 3.175
Male c .279 1.321 .045 1.006 1.736
Age .022 1.022 .004 1.007 1.038
______________________________________________________________________________
Note. a Compared to Not proctored. b Compared to African American. c Compared to Female.

Finally, results predicting retention demonstrated proctoring was not associated with

retention when controlling for the demographics. There is no relationship with proctoring. The

only significant finding in Table 6 shows Whites were more likely to be retained than African-

Americans, OR = 1.361 (CI = [1.047, 1.769]), p = .021.

 71

Table 6

Logistic Regression Statistics with Proctor and Demographics Predicting Retention

______________________________________________________________________________
95% CI for OR
Predictor B OR p LL UL

Proctor a .055 1.056 .698 .802 1.392

b
Race
Other races .134 1.144 .496 .777 1.683
White .308 1.361 .021 1.047 1.769
Male c -.081 .922 .553 .704 1.207
Age .004 1.004 .617 .989 1.019
______________________________________________________________________________
Note. a Compared to Not proctored. b Compared to African American. c Compared to Female.

Reliability and Validity

Ensuring there is reliability and validity in all research studies is important. Specifically,

there needs to be internal validity, external validity and objectivity. No statistical tests conducted

were researcher-conducted. Internal validity is the extent the researcher can conclude there is a

cause and effect relationship between the variables. The conclusion can only be a correct

inference if threats are accounted for in the study and design (Creswell, 2009). External validity,

as Creswell (2009) described, is the extent the researcher can determine the results of the study

can apply to a larger or extended population. If your research is applicable to other settings,

students, and classes, high validity would be validated. If the research cannot be replicated in

other situations, external validity is low. Scores used in research studies from participants

(students) are consistent over time are considered reliable (Creswell, 2009). Due to using

Inferential statistical tests have been used and determined to be valid and reliable over time

(AllPsych, 2018). Statistically proven tests were used with collected data from the PRE

department through college records of secondary data, and over 1900 students from five different
 72

mathematics courses, to uphold reliability and validity.

Chapter Summary

The purpose of the quantitative research was to examine whether proctoring assessments

during the semester were associated with final exam scores, course grades for success, and

retention for online mathematics students in the MACC area. The second purpose was to address

the three research questions and hypotheses addressed in the study. A total of 1909 cases across 5

courses from 2016 to 2018 were included in the study. Taken together, the study looked at

whether or not proctoring online math tests were significantly associated with final exam grades

and students’ success for the course.

Students who were proctored were less likely to pass the final exam and to succeed for the

course. Specifically, proctoring semester exams was 1.49 times more likely to fail the final exam

as compared to not proctoring semester exams. Worth noting, White and other races were more

likely to pass the final exam than African-Americans. Students who were proctored were also less

likely to succeed in the course. The result suggests students who attended the non-proctored

assessments during the semester were 1.53 times more likely to succeed in the course than the

student who had proctored assessments during the semester.

However, assessments being proctored was not related to whether or not a student was

retained in a math course. The only significant finding was Whites were more likely to be retained

than African-Americans. Therefore, the above findings reject the non-hypothesis for research

questions 1 and 2, and accept the non-hypothesis for research question 3. Chapter 5 analyses

findings of the study along with the study’s implications, limitations, and recommendations for

educators who are teaching online mathematics courses, and future researchers.
 73

Chapter 5: Discussion and Conclusion

The purpose of the quantitative research design is to examine the relationship of the

following final exam grades (pass/fail), percent of success (passing the semester with a 70% or

higher), and percent of retention of online mathematics students. Final exam scores, percent of

student success, and percent of student retention was be compared among classes with proctored

versus non-proctored tests during the semester. The comparison determined if there was a

relationship in online mathematics classes. The proposed study is necessary for the Mid-Atlantic

Community College (MACC) and the mathematics department, to determine if the college can

increase online success and decrease dropout rates through the type of assessments given to

students during the semester. Research indicates, online assessments are a challenge when

looking at a student’s performance (Hollister & Berenson, 2009). Studies have found the percent

of success in online classes are lower than in the comparable face-to-face classes (Borzewski,

2016). If the research is not conducted, there would not be a conclusion as to whether proctoring

tests affect the success of online students to meet or exceed those in face-to-face classes.

As seen in the literature review in Chapter 2, there is minimal research-based consensus

about the proposed study, which contributes to the knowledge base. The knowledge base helps

determine if proctored assessments affect successful education in online mathematics classes for

students in at MACC. To achieve the purpose of the study, three research questions were

addressed. Chapter 3 detailed the methods used to address the study’s research questions and

hypotheses. The quantitative design examined the relationship between a dichotomous variable

(“Proctored” or “Not Proctored”). Proctored was the independent variable while the dependent

variables were success rates, retention rates, and final exam scores. These were also broken up by

demographics.
 74

A generalized logistic mixed model (GLMM) was used to account for random effect of

Course ID and Term. There were five different math course used in the study, and the Course ID

was the name of the specific course, and the Term was the year and semester. Term indicates

time. Term is mainly used due to the duplicate cases in order to control for Term (same subject

took different courses for multiple terms).

In research question1, while considering gender, age, race, course ID, and term, students

who were proctored had higher rates of failing the final exam. White students pass rates for the

final exam were higher than African-Americans. Older participants final exam passing rates were

also higher. Gender and course ID were not associated with final grade.

In research question 2, while controlling for gender, age, race, course ID, and term,

students who were proctored were less likely to succeed. Whites had higher success rates than

African-American. Males showed higher scores than females. Older participants were more

successful than younger ones.

In research question 3, while controlling for gender, age, race, course ID, and term,

proctor was not associated with retention. There was no relationship with proctored tests and

retention. Whites had higher retention rates than African-American. gender, age, and course ID

were not associated with retention.

Chapter 5 includes a discussion of the findings in the study. Included is how the data may

be interpreted and conclusions which may be drawn from the results found. Chapter 5 also

discusses limitations in the study, recommendations for instructors of online mathematics courses

at community colleges, and future researchers. The chapter then presented the implications for

leadership at the community college level.

 75

Findings, Interpretations, and Conclusion

The results of the data analysis described in Chapter 4, give needed information to address

the study’s three research questions and hypotheses. Interpretations and conclusions from the

study can be determined within the context of the study’s theoretical framework. The outcomes

related to the research questions and hypotheses, in order to achieve the goals of the study, are

below.

Research Question 1: What is the relationship between student final exam scores,

(pass/fail) for students with proctored versus non-proctored online math tests?

Research Question 2: What is the relationship between the percent of student success for

proctored versus non-proctored online math tests?

Research Question 3: What is the relationship between the percent of student retention for

proctored versus non-proctored online math tests?

H10: There is no relationship between final exam scores (pass/fail) of students with

proctored versus non-proctored online tests.

H1A: There is a relationship between final exam scores (pass/fail) of students with

proctored versus non-proctored online tests.

H20: There is no relationship between the percent of student success with proctored versus

non-proctored tests.

H2A: There is a relationship between the percent of student success with proctored versus

non-proctored tests.

H30: There is no relationship between the percent of student retention with proctored

versus non-proctored tests.

 76

H3A: There is a relationship between the percent of student retention with proctored versus

non-proctored tests.

To address the hypothesis and research questions, generalized linear mixed models were

used to analyze the data. In research question 1, a participant was considered passing the final

exam when the grade was a 70% or above. The final exam grade was dichotomized into pass and

fail. Results revealed students who had proctored tests during the semester had higher rates of

failing the final exam. Older students along with White students had higher passing rates on the

final exam. Therefore, the above findings reject the non-hypothesis for research questions 1. H1A

is then true, stating there is a relationship between final exam scores (pass/fail) of students with

proctored versus non-proctored online tests.

In research question 2, the results suggest students who were assigned to non-proctored

assessments during the semester were 1.53 times more likely to succeed in the course. As with

success on final exam scores, Whites and older students had higher success rates in these online

mathematics course. To no surprise, students who do better on the final exam have higher success

rates in the class as well. The above findings reject the non-hypothesis for research questions 2.

H2A is then true, stating there is a relationship between the percent of student success with

proctored versus non-proctored tests.

In contrast, for research question 3, proctoring tests did not seem to determine if a student

was retained or not. When looking at race though, Whites were retained at a higher rate than

African-Americans. However, a greater proportion of participants who succeeded for the course

were retained (62.5%) than those who did not succeed (41.0%). Hence, the above findings accept

the non-hypothesis for research question 3. H30 is then true, stating there is no relationship

between the percent of student retention with proctored versus non-proctored tests.
 77

This research study is based on the theoretical framework of the Theory of Planned

Behavior, a cognitive theory, which aims to predict and understand the relationship between

human behavior and motivation (Peters, & Templin, 2010). Factors like motivation and

technology are impacting students cheating problem within institutions (Bolman & Deal, 2008).

Since the study determined students who were in online mathematics classes with non-proctored

tests, did better on final exams, and were more successful in the course, there may be future

studies needed. The studies may need to determine if the result is due to more students in the

proctored tests not showing up for the final exam because the student did not believe they were

going to be successful in the course. The study may also need to determine if due to the same

concept, more students in courses with non-proctored tests were more successful.

As noted in Chapter 2, there is conflicting information from studies when looking at

online courses. Demographics of who is being studied, could be influential on the research

studies. The study determined males had higher success rates than females. There are conflicting

results such as in the study done by White (2013), where overall, women were found to be more

successful than men in online courses. However, when determining who is successful on the final

exam or successful in the course for the semester, older students were determined to be the most

successful in the study. Students who are abreast in time management, multitasking, and critical

thinking, tend to take responsibility for personal learning, and are capable of working

independently, are typically adult learners. The students are usually more mature and possess

prior knowledge and life experiences and can relate to education and have effectively introduced

these elements back into online education classes (Johnson & Berge, 2012).

Demographics can affect retention, according to Elam (2013). Students who are educated

on how these classes work, operate, and are organized, possess the ability to improve retention
 78

and success (Elam, 2013). Research Question 3 found proctoring tests did not have an effect on

retention. However, White students were retained more often than African-American students.

Future research could be conducted to determine if there is a connection between demographics at

the MACC and student education in online mathematics courses.

The data analyzed in the study showed Whites had higher success rates on online

mathematics final exams, in passing an online mathematics course, and being retained. Students

in non-proctored online mathematics courses had higher success rates when looking at final

exams and passing the course for the semester. Proctored assessments on the other hand, did not

seem to have an effect on retention either way. The study contributed additional data to the

growing body of research exploring the relationship between proctored and non-proctored tests in

online mathematics courses at the community college level.

Limitations

Lurking variables having the potential to impact the student’s performance within the

research sample may have existed. Other demographic variables which could have a potential

impact on students’ performance include age, having to take care of family members, and the

number of hours students need to work outside of the education commitment (Glazier, 2016). The

research sample was diverse in these and students of all demographics in the study had equal

opportunity to sign up for an online mathematics course in place of a traditional face-to-face

course.

A three-year period has been selected for the focused research to include a variety of

instructors, both full time and adjunct, as well as a variety and diverse population of students.

During the given three year time frame, the online classes in the study were institutionalized,

which means each course was set up the same, and the assessments and homework’s throughout
 79

the semester were the same with only random numbers of the same difficulty changing in each

problem. Common courses gave the students as close to the same experience in each class as

possible. These students were in five different levels of math to help address any potential student

or instructor bias. The scope of the study is generalizable because of the population the study was

pulled from a large diverse population of 62,000 students with about 29% Pell Grant recipients,

45% of the students body is white, and 33% African American, 6% Hispanic or Latino, 6% Asian,

2% multicultural, and 8% unknown. Statistics here are comparable to the findings to other online

mathematics courses at the MACC and possibly other community colleges with similar

demographics.

The study was limited to five online mathematics courses which included Pre-Algebra,

Introductory Algebra, Intermediate Algebra, College Algebra, and Pre-Calculus. Other online

math courses offered at MACC were not used in the study due to lack of institutionalized courses

which lead to many other variables which could contribute to results in the study. These courses

can use other online homework and assessment systems, have individually created tests at

multiply difficulty levels, and more.

Recommendations

Instructors and administrators at community colleges continue to be given the

responsibility of setting appropriate requirements and standards for online mathematics courses.

The literature review in Chapter 2 showed the need for additional studies on the relationship

between proctored and non-proctored assessments in online mathematics courses as related to

success. Further research is recommended at the MACC as well as in other diverse settings

throughout the United States.

 80

Administration and leaders at the institution should consider further research as more

online mathematics courses get institutionalized. Another study may be beneficial is to see how

many students took the final exam in proctored settings versus non-proctored settings to

determine if there is a possible relationship in success rates. A point of interest may also be to

note how many students drop or no longer participate in proctored and non-proctored classes to

compare the relationship.

This study contained about 1900 cases over three years and five different mathematics

courses over a diverse population. Potentially the findings can be generalized to other community

colleges with similar populations as the study. Future researchers may decide to consider different

methodologies in order to address the relationship between proctored and non-proctored

assessments in online math courses. For instance, the study did not include any qualitative data

from students or instructors to better understand success and retention. Including qualitative data

or mixed-methods study may serve to explain student performance and success characteristics

found to be important by administration and instructors. Included qualitative interviews or

surveys may provide additional insight into perceptions about the reasons for success and

retention rates.

Implications for Leadership

This study is significant to administration, educators, and instructors, as results are

revealed for proctoring and not proctoring tests in online mathematics courses in reference to

success and retention. There was evidence a relationship existed between students in classes with

non-proctored assessments during the semester and being more successful on final exams and the

course as a whole. As for retention, more studies have to be completed to determine what

variables are effecting students and determining if students are going to sign up for the
 81

subsequent math course or not.

Leadership can now determine based on the study, what is the most beneficial way to

conduct online courses in the mathematics department. The study also leaves the door open for

future research to be conducted to determine possible reasons for these conclusions. The

community college administrators have looked into multiple online proctoring methods and

taking the final exams for math classes with paper and pencil for instructors to see all algebraic

work. As variables change, so may relationships, successes, and outcomes.

Conclusions

Chapter 5 provided an overview of some of the previous chapters in the study. The

problem researched, the purpose of the study, and the methods used to answer and summarize

research questions and hypotheses. The study was to be determined whether or not proctoring

online math tests were significantly associated with final exam grade and students’ success for the

course. Students who were given proctored assessments during the semester were less likely to

pass the final exam and to succeed for the course. However, proctor was not related to whether or

not a student retained in math course. Therefore, the above findings reject the non-hypothesis for

research questions 1 and 2, and accept the non-hypothesis for research question 3. These

conclusions can potentially be assumed for other institutionalized online math courses at MACC

and other diverse community colleges.

Chapter 5 discussed the limitations of the study according to the study’s research

questions, the existing literature from Chapter 2, and the theoretical framework directing the

study. Future research may focus on variables which cause retention, and the number of cases

who may drop or no longer attend proctored classes and are given a 0% for the final exam and

hence fail the course. Because community colleges continue to be tasked with setting appropriate
 82

requirements and standards in online courses, further research examining relationships with

proctored assessments was recommended. Chapter 5 also provided recommendations for future

researchers choosing to contribute to the growing body of research and knowledge in the field. A

better understanding of the relationship between proctored tests and non-proctored tests can now

be understood.
 83

References

Adinoff, B., Conley, R. R., Taylor, S. F., & Chezem, L. L. (2013). Protecting confidentiality in

human research. American Journal of Psychiatry, 170(5), 466-470. Retrieved from

https://ajp.psychiatryonline.org/

Al-Dossary, S. (2017). Why do college students cheat? A structural equation modeling validation

of the theory of planned behavior. International Education Studies, 10(8), 40–51.

Retrieved from http://www.ccsenet.org/journal/index.php/ies

Alessio, H. M., Malay, N., Maurer, K., Bailer, A. J., & Rubin, B. (2017). Examining the effect of

proctoring on online test scores. Online Learning, 21(1), 146–161. Retrieved from

https://olj.onlinelearningconsortium.org/index.php/olj

Allen, I., & Seaman, J. (2014). Grade change: Tracking online education in the United States.

Babson Park, MA: Babson Survey Research Group, Babson College. Retrieved from

https://www.onlinelearningsurvey.com/reports/gradechange.pdf

Allen, I., Seaman, J., Babson Survey Research Group, & Quahog Research Group. (2015). Grade

level: Tracking online education in the United States. Babson Survey Research Group.

Retrieved from https://www.babson.edu/academics/faculty/provost/babson-survey-

research-group/

AllPsych. (2018). Variables, validity and reliability. Retrieved from

https://allpsych.com/researchmethods/

Anderson, S., Kelley, K., & Maxwekk, S. (2017). Sample-size planning for maore accurate

statistical power: A method adjusting sample effect sizes for publication bias and

uncertainty. Psychological Science, 28(11), 1547–1562. Retrieved from

https://doi.org/10.1177/0956797617723724
 84

Ardid, M., Gomez-Teiedor, J., Meseguer-Duenas, J., Riera, J., & Vidaurre, A. (2015). Online

exams for blended assessment. Study of different application methodologies. Computers

& Education, 81(1), 296–303. doi:10.1016/j.compedu.2014.10.010

Ashby, J., Sadera, W. A., & McNary, S. W. (2011). Comparing student success between

developmental math courses offered online, blended, and face-to-face. Journal of

Interactive Online Learning, 10(3), 128–140. Retrieved from http://www.ncolr.org/

Ayal, S., Gino, F., Barkan, R., & Ariely, D. (2015). Three principles to revise people’s unethical

behavior. Perspectives on Psychological Science, 10(6), 738–741. Retrieved from

https://journals.sagepub.com/home/pps

Bailey, A., Barton, C., & Mullen, K. (2014). The five faces of online education: What students

and parents want. BCG Perspectives, 3-18. Retrieved from http://image-

src.bcg.com/Images/Five_Faces_Online_Education_Jun_2014_tcm9-61505.

Bain, L. E. (2017). Rethinking research ethics committees in low-and medium-income countries.

Retrieved from https://doi.org/10.1177/1747016117692026

Beck, H. P., & Milligan, M. (2014). Factors influencing the institutional commitment of online

students. The Internet and Higher Education, 20, 51–56.

doi:10.1016/j.iheduc.2013.09.002

Bernard, R., Abrami, P., Borokhovski, E., Wake, C., Tamim, R., & Surkes, M. (2014). A meta-

analysis of three types of interaction treatments in distance education. Review of

Educational Research, 79(3), 1243–1289. Retrieved from

http://www.aera.net/Publications/Journals/Review-of-Educational-Research

Best, S. (2014). Understanding and doing successful research: Data collection and analysis. New

York, NY: Pearson Education Limited.

 85

Betts, K., Welsh, B., Pruitt, C., Hermann, K., Dietrich, G., Trevino, J. G., & Coombs, N. (2013).

Understanding disabilities & online student success. Journal of Asynchronous Learning

Networks, 17(3), 15–48. Retrieved from https://onlinelearningconsortium.org/read/journal-

issues/

Blachnio, A., & Weremko, M. (2011). Academic cheating is contagious: The influence of the

presence of others on honesty. International Journal of Applied Psychology, 1(1), 14–

19. Retrieved from http://www.sapub.org/journal/aimsandscope.aspx?journalid=1027

Black, R. C. (2018). Critical assessment and strategies for increased student retention. Hershey,

PA: Information Science Reference.

Bloemers, W., Oud, A., & van Dam, K. (2016). Cheating on unproctored Internet intelligence

tests: Strategies and effects. Personnel Assessment and Decisions, 2, 21–29.

https://doi.org/ 10.25035/pad.2016.003

Blue, A. (2018). Exploring mentoring strategies needed by higher educational managers to

increase student retention rates and decrease dropout rates in a higher education

organization. Ann Arbor, MI: ProQuest LLC.

Bolman, L., & Deal, T. (2008). Reframing organizations: Artistry, choice, and leadership. San

Francisco, CA: Jossey Bass.

Borzewski, L. M. (2016). Factors for success in community college online gateway math.

(Doctoral dissertation). Retrieved from Walden Dissertations and Doctoral Studies.

Boslaugh, S. E. (2013). Theory of planned behavior. Salem Press Encyclopedia. Retrieved from

https://www.salempress.com/
 86

Boston University. (2018). The theory of planned behavior. Boston University School of Public

Health. Retrieved from http://sphweb.bumc.bu.edu/otlt/MPH-

Modules/SB/BehavioralChangeTheories/BehavioralChangeTheories3.html

Brallier, S., & Palm, L. (2015). Proctored and unproctored test performance. International

Journal of Teaching & Learning in Higher Education, 27(2), 221–226. Retrieved from

http://www.isetl.org/ijtlhe/

Carstairs, J., & Myors, B. (2009). Internet testing: A natural experiment reveals test score

inflation on a high-stakes, unproctored cognitive test. Computers in Human Behavior, 25,

738–742. https://doi.org/10.1016/j.chb.2009.01.011

Center for Innovation in Mathematics. (2011). An analysis of how proctoring exams in online

mathematics offerings affects student learning and course integrity. Retrieved from

https://docplayer.net/4849139-An-analysis-of-how-proctoring-exams-in-online-

mathematics-offerings-affects-student-learning-and-course-integrity.html, 1-13.

Child Trends Databank. (2015). Family structure. Retrieved from http://www.childtrends.org/?

indicators=familystructure

Chu, S., Chen, H., & Sung, Y. (2016). Following brands on Twitter: An extension of theory of

planned behavior. International Journal of Advertising, 35, 421–437. https://doi.org/ch66

Cifuentes, L., & Janney, A. (2016). Protecting students’ integrity and reducing academic

dishonesty in online learning. Distance Learning, 13(4), 9–15. Retrieved from

https://www.infoagepub.com/distance-learning

Cochran, J. J., Campbell, S., Baker, H., & Leeds, E. (2014). The role of student characteristics in

predicting retention in online courses. Research in Higher Education, 55(1), 27–48.

doi:10.1007/s11162-013-9305-8
 87

Cohen, L., Lawrence, M., & Morrison, K. (2017). Research methods in education. New York,

NY: Routledge.

Community College of Baltimore County. (2018). Community college of Baltimore county.

Retrieved from www.ccbcmd.edu.

Cotrell, R., & McKenzie, F. (2011). Health promotion & education research methods. Sudbury,

MA: Jones and Bartlett Publishers.

Creswell, J. W. (2005). Educational research: Planning, conducting, and evaluating quantitative

and qualitative research. Upper Saddle River, NJ: Pearson Education, Inc.

Creswell, J. W. (2009). Research design: Qualitative, quantitative, and mixed methods

approaches. Los Angeles, CA: Sage.

Creswell, J. W. (2017). Research design: Qualitative, quantitative, and mixed methods

approaches. Thousand Oaks, CA: Sage.

Cuesta, H., & Kumar, S. (2016). Practical data analysis. Birmingham, UK: Packt Publishing.

Curtis, G., Cowcher, E., Greene, B., Rundle, K., Paull, M., & Davis, M. (2018). Self-control,

injunctive norms, and descriptive norms predict engagement in plagiarism in a theory of

planned behavior model. Journal of Academic Ethics, 16(3), 225–239.

https://doi.org/10.1007/s10805-018-9309-2

Daffin, L. W., & Jones, A. A. (2018). Comparing student performance on proctored and non-

proctored exams in online psychology courses. Online Learning, 22(1), 131-145.

doi:10.24059/olj.v22i1.1079

Deshpande, B. (2011). Two key assumptions to be aware of before applying the chi-square test.

Retrieved April 13, 2019, from http://www.simafore.com/blog/bid/56480/2-key-

assumptions-to-be-aware-of-before-applying-the-chi-square-test.
 88

DeSimone, J., Harms, P., & DeSimone, A. (2015). Best practice recommendations for data

screening. Journal of Organizational Behavior, 36(2), 171-181. Retrieved from

http://digitalcommons.enl.edu/managementfacpub/124

Diedenhofen, B., & Musch, J. (2017). PageFocus: Using paradata to detect and prevent cheating

on online achievement tests. Behavior Research Methods, 49, 1444–1459. https://doi.org/

10.3758/s13428-016-0800-7

Eisenberg, J. (2004). To cheat or not to cheat: Effects of moral perspective and situational

variables on students’ attitudes. Journal of Moral Education, 33(2), 163–177. Retrieved

from https://www.tandfonline.com/loi/cjme20

Elam, E. (2013). A quantitative inquiry into the factors influence success in online classes.

Retrieved from https://www.learntechlib.org/p/115500/.

Enago Academy (2019). Importance of the results section. Retrieved from

https://www.enago.com/academy/importance-of-results-section-in-academic-papers/

Farnesea, M., Tramontanob, C., Fidaa, R., & Pacielloc, M. (2011). Cheating behaviors in

academic context: Does academic moral disengagement matter? Procedia - Social and

Behavioral Sciences, 29, 356–365. Retrieved from

https://www.journals.elsevier.com/procedia-social-and-behavioral-sciences/

Fask, A., Englander, F., & Wang, Z. (2014). Do online exams facilitate cheating? An experiment

designed to separate possible cheating from the effect of the online test taking

environment. Journal of Academic Ethics, 12(2), 101–112. Retrieved from

https://link.springer.com/journal/10805
 89

Fask, A., Englander, F., & Wang, Z. (2015). The integrity of online testing for introductory

statistics courses: A latent variable approach. Practical Assessment, Research &

Evaluation, 20(8–12), 1–12. Retrieved from https://pareonline.net/

Ferrer, D. (2019). History of online education. Retrieved from

https://thebestschools.org/magazine/online-education-history/

Flowers, L. O., White, E. N., Raynor, J. E., & Bhattacharya, S. (2012). African American

students’ participation in online distance education in STEM disciplines: Implications for

HBCUs. Sage Open Journals, 2(2), 1–5. doi:10.1177/2158244012443544

Friedman, L. W., & Friedman, H. H. (2013). Crises in education: Online learning as a solution.

Creative Education, 2, 156–163. Retrieved from https://www.scirp.org/journal/ce/

Glazier, R. A. (2016). Building rapport to improve retention and success in online classes.

Journal of Political Science Education, 12(4), 437–456.

https://doi.org/10.1080/15512169.2016.1155994

Gomez, G. (2013). The effects of retention as public policy in Latin America. Retrieved from

https://www.researchgate.net/publication/253644433_Gomez_G_2013_The_effects_of_re

tention_as_public_policy_in_Latin-

America_XV_Comparative_Education_World_Congress_WCCES2013_Junio_24_-

_28_2013_Buenos_Aires_Argentina

Gosling, S. D., Sandy, C. J., John, O. P., & Potter, J. (2010). Wired but not weird: The promise of

the internet in reaching more diverse samples. Behavioral and Brain Sciences, 33(2–3),

94. Retrieved from https://www.cambridge.org/core/journals/behavioral-and-brain-

sciences
 90

Gregory, C. B., & Lampley, J. H. (2016). Community college student success in online versus

equivalent face-to face courses. Journal of Learning in Higher Education, 12(2), 63–72.

Retrieved from http://www.jwpress.com

Gul, S., Shafiq, H., Mahajan, I., Shafi, M., & Shah, T. A. (2018). Massive open online courses:

Hype and hope. DESIDOC Journal of Library & Information Technology, 38(1), 63–66.

https://doi.org/10.14429/djlit.38.1.11141

Haigh, M. (2006). Promoting environmental education for sustainable development: The value of

links between higher education and non-governmental organizations. Journal of

Geography in Higher Education, 30 (2), 327–349.

Harari, G. M., Lane, N. D., Wang, R., Crosier, B. S., Campbell, A. T., & Gosling, S. D. (2016).

Using smartphones to collect behavioral data in psychological science: Opportunities,

practical considerations, and challenges. Perspectives on Psychological Science: A

Journal of the Association for Psychological Science, 11(6), 838–854. Retrieved from

https://www.psychologicalscience.org/publications/perspectives

Harmon, O., & Lambrinos, J. (2008). Are online exams an invitation to cheat? Journal of

Economic Education, 39(2), 116-125. Retrieved from

https://www.tandfonline.com/loi/vece20

Haynie, D. (2015). Younger students increasingly drawn to online learning, study finds. Retrieved

from https://www.usnews.com/education/online-education/articles/2015/07/17/younger-

students-increasingly-drawn-to-online-learning-study-finds

Hollister, K., & Berenson, M. (2009). Proctored versus unproctored online exams: Studying the

impact of exam environment on student performance. Decision Sciences Journal of

Innovative Education, 7(1), 271–294. doi:10.1111/j.1540-4609.2008.00220.x

 91

Houser, R. (2015). Counseling and educational research: Evaluation and application. Thousand

Oaks, CA: Sage Publications.

Hoy, W., & Adams, C. (2016). Quantitatie research in education: A primer. Thousand Oaks, CA:

SAGE.

Ihme, J. M., Lemke, F., Lieder, K., Martin, F., Müller, J. C., & Schmidt, S. (2009). Comparison of

ability tests administered online and in the laboratory. Behavior Research Methods, 41,

1183–1189. https://doi.org/10.3758/BRM.41.4.1183

Jackson, S. (2015). Research methods and statistics: A critical thinking approach. New York,

NY: Cengage Learning.

Jaggars, S. S., & Xu, D. (2010). Online learning in the Virginia community college system.

Community College Research Center, Columbia University. Retrieved from

https://ccrc.tc.columbia.edu/publications/online-learning-virginia.html

James, S., Swan, K., & Daston, C. (2016). Retention, progression and the taking of online

courses. Online Learning, 20(2), 189–210. Retrieved from

https://olj.onlinelearningconsortium.org/index.php/olj

Jayaratne, K., & Moore, G. (2017). Perceptions of college students toward online classes:

Implications for teaching online. NACTA Journal, 61(4), 304–309. Retrieved from

https://www.nactateachers.org/index.php/journal-sp-1148215168

Jensen, C., & Thomsen, J. (2014). Self-reported cheating in web surveys on political knowledge.

Retrieved from doi: 10.1007/s11135-013-9960-z

Johnson, H., Cuellar Mejia, M., & Cook, K. (2015). Successful online courses in California’s

community colleges. San Francisco, CA: Public Policy Institute of California.

 92

Johnson, S., & Berge, Z. (2012). Online education in the community college. Community College

Journal of Research and Practice, 36(11), 897–902. doi:10.1080/10668920903323948.

Kalinski, F. (2015). Transforming student retention in higher education online programs in

California community colleges: A Delphi study. Retrieved from

https://digitalcommons.brandman.edu/cgi/viewcontent.cgi?article=1103&context=edd_dis

sertations.

Karim, M., Kaminsky, S., & Behrend, T. (2014). Cheating, reactions, and performance in

remotely proctored testing: An exploratory experimental study. Journal of Business &

Psychology, 29(4), 555–572. Retrieved from http://www.jstor.org/stable/24709885

Khemakhem, S. (2016). Investigating the predictive validity of IELTS for a teacher education

program in UAE. Retrieved from http://eprints.uwe.ac.uk/28078/

Kirkman, C. J., McNees, H., Stickl, J., Banner, J. H., & Hewitt, K. K. (2016). Crossing the

suspension bridge: Navigating the road from school suspension to college success—How

some students have overcome the negative implications of school suspension to bridge the

road to college. Journal of Organizational and Educational Leadership, 2(1). Retrieved

from http://digitalcommons.gardner-webb.edu/joel/

Kolowich, S. (2013). Behind the webcam’s watchful eye, online proctoring takes hold. The

Chronicle of Higher Education, 59(32), 12. Retrieved from

https://chronicle.com/article/Behind-the-Webcams-Watch ful/138505/

Kolski, T., & Weible, J. (2018). Examining the relationship between student test anxiety and

webcam based exam proctoring. Online Journal of Distance Learning

Administration, 21(3), 1-15. Retrieved from https://www.westga.edu/~distance/ojdla/

 93

Kovacs, P., Peslak, A., Kovalchick, L., Wang, W., & Davis, G. A. (2017). Effectiveness of course

delivery methods: A multi-university study. Issues in Information Systems, 18(1), 137.

Retrieved from https://www.iacis.org/iis/iis.php

Ladyshewsky, R. K. (2015). Post-graduate student performance in “supervised in-class” vs.

“unsupervised online” multiple choice tests: Implications for cheating and test security.

Assessment & Evaluation in Higher Education, 40(7), 883–897.

doi:10.1080/02602938.2014.956683

Leavy, P. (2017). Research design: Quantitative, qualitative, mixed methods, arts-based, and

community-based research approaches. New York, NY: Guilford Press.

Lederman, D. (2018). Who is studying online (and where). Retrieved from

https://www.insidehighered.com/digital-learning/article/2018/01/05/new-us-data-show-

continued-growth-college-students-studying

Lee, Y., Stringer, D., & Jianjun, D. (2017). What determines students’ preference of online to

F2F class? Business Education Innovation Journal, 9(2), 97–102. Retrieved from

http://www.beijournal.com/

Levesque, E. (2018). Improving community college completion rates by addressing structural and

motivational bariers. Retrieved from https://www.brookings.edu/research/community-

college-completion-rates-structural-and-motivational-barriers/

MacInnes, J. (2017) An introduction to secondary data analysis with IBM SPSS statistics.

London, UK: SAGE.

Madara, D., Namango, S., & Katana, H. (2016). Theories and models relevant to cheating-

behaviour. Research on Humanities and Social Sciences, 6(17), 10–138. Retrieved from

https://iiste.org/Journals/index.php/RHSS/index
 94

Martin, F., Wang, C., & Sadaf, A. (2018). Student perception of helpfulness of facilitation

strategies that enhance instructor presence, connectedness, engagement and learning in

online courses. The Internet and Higher Education, 3(7), 52–

65.doi:10.1016/j.iheduc.2018.01.003

Matika, R. S. (2012). Student and instructor perceptions of factors important for students’ success

in online and in-person algebra classes at Somerset Community College (Doctoral

dissertation). Retrieved from UKnowledge.

McGinn, M. (2018). Teaching and researching ethically: Guidance for instructor-researchers,

educational developers, and research ethics personnel. Canadian Journal for the

Scholarship of Teaching and Learning , 9(1), 1-14. Retrieved from

https://doi.org/10.5206/cjsotl-rcacea.2018.1.2

McQuillan, P., & Zito, N. (2011). It’s not my fault: Using neutralization theory to understand

cheating by middle school students. Current Issues in Education, 13(3), 1–25. ISSN 1099-

839X

Means, B., Toyama, Y., Murphy, R., Bakia, M., & Janes, K. (2009). Evaluation of evidence-based

practices in online learning: A meta-analysis and review of online learning studies. U.S.

Department of Education, 1-66. Retrieved from

http://www2.ed.gov/rschstat/eval/tech/evidence-based-practices/finalreport.pdf

Millea, M., Wills, R., Elder, A., & Molina, D. (2018). What matters in college student success?

Determinants of college retention and graduation rates. Education, 138(4), 309-322.

Miller, A., & Young-Jones, A. D. (2012). Academic integrity: Online classes compared to face-

to-face classes. Journal of Instructional Psychology, 39(3–4), 138-140. Retrieved from

https://www.questia.com/library/p6137/journal-of-instructional-psychology
 95

Miller, G. (2012). The smartphone psychology manifesto. Perspectives on Psychological Science,

7, 221–237. https://doi.org/10.1177/1745691612441215

Mitchel, S. (2017). Online learning: No worries at community colleges. HETS Online Journal, 7,

5–20. Retrieved from https://hets.org/tag/hets-online-journal/

Moore, D. (2014). An investigation of the attrition of African-American students in an online

undergraduate program (Doctoral dissertation). Retrieved from NOVA Southeastern

University.

Moore, J. C., & Fetzner, M. J. (2009). The road to retention: A closer look at institutions that

achieve high course completion rates. Journal of Asynchronous Learning Networks, 13(3),

3–22. Retrieved from https://secure.onlinelearningconsortium.org/publications/olj_main

Moore, P. H., Head, J. D., & Griffin, R. B. (2017). Impeding students’ efforts to cheat in online

classes. Journal of Learning in Higher Education, 13(1), 9–24. Retrieved from

jwpress.com/JLHE/JLHE.htm

Moten, J., Jr., Fitterer, A., Brazier, E., Leonard, J., & Brown, A. (2013). Examining online college

cyber cheating methods and prevention measures. The Electronic Journal of eLearning,

11, 139–146. Retrieved from http://www.ejel.org/main.html

Mozes-Carmel, A., & Gold, S. S. (2009). A comparison of online vs. proctored final exams in

online classes. Journal of Educational Technology, 6(1), 76–81. Retrieved from

https://journals.sagepub.com/home/ets

Murdock, T., Hale, N., & Weber, M. (2001). Predictors of cheating among early adolescents:

Academic and social motivations. Contemporary Educational Psychology, 26, 96–115.

Retrieved from https://www.journals.elsevier.com/contemporary-educational-psychology

 96

Murphy, C. A., & Stewart, J. C. (2017). On-campus students taking online courses: Factors

associated with unsuccessful course completion. The Internet and Higher Education, 34,

1–9. doi:10.1016/j.iheduc.2017.03.001

MyOpenMath. (2018). MyOpenMath. Retrieved from www.myopenmath.com

Nash, J. (2015). Future of online education in crisis: A call to action. The Turkish Online Journal

of Educational Technology. 14(2). 80-88. Retrieved from http://www.tojet.net/

National Center for Educational Statistics, (2015). School composition and the black-white

achievement gap. Retrieved from

https://nces.ed.gov/nationsreportcard/subject/studies/pdf/school_composition_and_the_bw

_achievement_gap_2015.pdf

Nese, J., Lai, C., & Anderson, D. (2013). A primer on longitudinal data analysis in education.

Behavioral Research and Teaching. Retrieved from

http://dx.doi.org/10.5430/ijhe.v4n3p188

Neuman, W. L. (2003). Social research methods: Qualitative and quantitative approaches.

Boston, MA: Pearson Education, Inc.

O'Connell, M. (2018). Proctored vs. unproctored testing: Does it really make a Difference?

Retrieved from http://www.selectinternational.com/proctored-vs.-unproctored-testing-

does-it-really-make-a-difference

Online Learning Consortium. (2015). Five pillars of quality online education. Retrieved from:

http://onlinelearningconsortium.org/about/quality-framework-five-pillars/

Ouzts, K. (2006). Sense of community in online courses. Quarterly Review of Distance

Education, 7(3), 285–296. Retrieved from https://www.infoagepub.com/quarterly-review-

of-distance-education
 97

Ozerbas, M. O., & Erdogan, B. B. (2016). The effect of the digital classroom on academic

success and online technologies self-efficacy. Journal of Educational Technology &

Society, 19(4), 203–212. Retrieved from http://www.ifets.info

Papia, B. (2016). Retention in online courses. SAGE Open, 1(1), 1-11.

doi:10.1177/2158244015621777

Peersman, G. (2014). Data collection and analysis methods in impact evaluation: Methodological

briefs. Retrieved from https://www.unicef-irc.org/publications/755-overview-data-

collection-and-analysis-methods-in-impact-evaluation-methodological.html

Peslak, A., Kovalchick, L., Wang, W., & Kovacs, P. (2018). Attitudes toward course delivery: A

multi-university study of online, on-ground, and hybrid instruction. Information Systems

Education Journal, 16(4), 27–33. Retrieved from http://isedj.org/

Peters, R., & Templin, T. (2010). Theory of planned behavior, self-care motivation, and blood

pressure self-care. Research and Theory for Nursing Practice, 24(3), 172–86. Retrieved

from https://www.springerpub.com/research-and-theory-for-nursing-practice.html

Platt, C. A., Raile, A., & Yu, N. (2014). Virtually the same?: Student perceptions of the

equivalence of online classes to face-to-face classes. Journal of Online Learning &

Teaching, 10(3), 489–503. Retrieved from http://jolt.merlot.org/

Quality Matters. (2018). Helping you deliver your online promise. Retrieved from

https://www.qualitymatters.org/

Reddy, M., & Andrade, H. (2010). A review of rubric use in higher education, assessment &

evaluation in higher education, 35:4, 435-448, DOI: 10.1080/02602930902862859

 98

Rey, J. G. (2010). The effects of online courses for student success in basic skills mathematics

classes at California community colleges. ProQuest LLC, 1-95. Retrieved from

http://digitallibrary.usc.edu/cdm/ref/collection/p15799coll127/id/332845

Richardson, A. S. (2018). A study of the relationship among revenue sources and retention rates

at private historically black colleges and universities. Ann Arbor, MI: ProQuest LLC.

Retrieved from https://www.ncat.edu/

Salvo, S., Shelton, K., & Welch, B. (2017). African American males and online education: A

review of the literature. Online Journal of Distance Learning Administration, 20(4), 1-20.

Retrieved from https://www.westga.edu/~distance/ojdla/

Samavati, H., Stumph, C. F., & Dilts, D. A. (2012). Interesting times: Relation between test times

and student performance in online courses. Journal of Higher Education Theory &

Practice, 12(6), 74–80. Retrieved from http://www.na-businesspress.com/jhetpopen.html

Sanford, D., Ross, D., Rosenbloom, A., & Singer, D. (2017). Course convenience, perceived

learning, and course satisfaction across course formats. E-Journal of Business Education

& Scholarship of Teaching, 11(1), 69–84. Retrieved from

https://cetl.kennesaw.edu/journals/e-journal-business-education-and-scholarship-teaching

Scott, G. A. (2013). Status of the department of education’s inventory of its data collections.

Retrieved from https://www.gao.gov/assets/660/655668.pdf

Shea, P., & Bidjerano, T. (2014). Does online learning impede degree completion? A national

study of community college students. Computers & Education, 75, 103–111. Retrieved

from https://www.journals.elsevier.com/computers-and-education
 99

Shotwell, M., & Apigian, C. H. (2015). Student performance and success factors in learning

business statistics in online vs. on-ground classes using a web-based assessment platform.

Journal of Statistics Education, 23(1), 1-19. http://www.tandfonline.com

Spivey, M. F., & McMillan, J. J. (2014). Classroom versus online assessment. Journal of

Education for Business, 89(8), 450–456.

https://www.tandfonline.com/toc/vjeb20/89/8?nav=tocList

Stack, S. (2015). The impact of exam environments on student test scores in online courses.

Journal of Criminal Justice Education, 26(3), 273–282. Retrieved from

https://www.tandfonline.com/loi/rcje20

Statistics Solutions. (2013). Data analysis plan: Chi-square test of independence. Retrieved from

http://www.statisticssolutions.com/academic-solutions/member-resources/member-

profile/data-analysis-plan-templates/data-analysis-plan-chi-square-test-of-independence

Steger, D., Schroeders, U., & Gnambs, T. (2018). A meta-analysis of test scores in proctored and

unproctored ability assessments. European Journal of Psychological Assessment.

https://doi.org/10.1027/1015-5759/a000494

Stowell, J. R. (2015). Online open-book testing in face-to-face classes. Scholarship of Teaching

and Learning in Psychology, 1(1), 7–13. Retrieved from

https://www.apa.org/pubs/journals/stl/

Struble, K. (2014). Efficacy of hybrid coursework on retention rate in online higher education.

Retrieved from https://digitalcommons.liberty.edu/doctoral/970/

Swartz, L. B., & Cole, M. T. (2013). Students’ perception of academic integrity in online business

education courses. Journal of Business & Educational Leadership, 4(1), 102–112.

 100

Retrieved from https://www.questia.com/library/p439713/journal-of-business-and-

educational-leadership

Sweet, S., & Grace-Martin, K. (2012). Data Analysis with SPSS: A First Course in Applied

Statistics. Boston, MA: Allyn & Bacon.

Tao, J., & Li, X. (2012). A case study on computerized take-home testing: Benefits and pitfalls.

International Journal of Technology in Teaching and Learning, 8(1), 33–43. Retrieved

from http://stelar.edc.org/publisher/international-journal-technology-teaching-and-learning

Tendeiro, J. N., Meijer, R. R., Schakel, L., & Maij-de Meij, A. M. (2013). Using cumulative sum

statistics to detect inconsistencies in unproctored Internet testing. Educational and

Psychological Measurement, 73, 143–161. https://doi.org/10.1177/ 0013164412444787

Travers, S. (2016). Supporting online student retention in community colleges. Quarterly Review

of Distance Education, 17(4), 49–61. Retrieved from

https://www.infoagepub.com/quarterly-review-of-distance-education

Trenholm, S. (2009). A study on the efficacy of computer-mediated developmental math

instruction for traditional community college students. Research & Teaching in

Developmental Education, 25(2), 68–76. Retrieved from

https://www.jstor.org/journal/reseteacdeveeduc

Tucker, W. G. (2014). Spaces for success in higher education: Males of color at an online

predominantly white community college (Doctoral dissertation). Retrieved from Northern

Arizona University.

United States Census Bureau. (2015). Table C8: Poverty status, food stamp receipt, and public

assistance for children under 18 years by selected characteristics. Retrieved from

 101

https://www2.census.gov/programssurveys/demo/tables/families/2015/cps2015/tabc8-

all.xls

University of Connecticut. (2018). Educational research basics.

https://researchbasics.education.uconn.edu/variables/#

University of Leeds. (2018). Literature searching explained. Retrieved from

https://library.leeds.ac.uk/info/1404/literature_searching/14/literature_searching_explaine

d/4

University of Pennsylvania School of Arts & Sciences, (2008). Tutorial: Pearson’s Chi-square test

for independence. Retrieved from https://www.ling.upenn.edu/~clight/chisquared.htm.

U.S. Department of Education. (2014). Digest of education statistics. Retrieved from

https://nces.ed.gov/programs/digest/d15/tables/dt15_311.15.asp.

U.S. Department of Education. (2015). Digest of education statistics. Retrieved from

https://nces.ed.gov/fastfacts/display.asp?id=80.

U.S. Department of Education. (2015a). Fact sheet: Focusing higher education on student success.

Retrieved from https://www.ed.gov/news/press-releases/fact-sheet-focusing-higher-

education-student-success

U.S. Department of Education. (2017). Enrollment and employees in postsecondary institutions,

fall 2016; and financial statistics and academic libraries, fiscal year 2016. Retrieved from

https://nces.ed.gov/pubs2018/2018002.pdf

Varble, D. L., & Haute, T. (2014). Reducing cheating opportunities in online tests. Atlantic

Marketing Journal, 3(3), 131–149. Retrieved from

https://digitalcommons.kennesaw.edu/amj/
 102

Wagner, E., Enders, J., Pirie, M., & Thomas, D. (2016). Supporting academic integrity in a fully-

online degree completion program through the use of synchronous video conferences.

Journal of Information Systems Education, 27(3), 159. Retrieved from http://jise.org/

Waruingi, M. (2013). Dr. Mac! Dissertation mentoring handbook: Book 1: Strategies for

quantitative research. Morrisville, NC: Lulu Press.

Watson, G., & Sottile, J. (2015). Cheating in the digital age: Do students cheat more in online

courses? Online Journal of Distance Learning Administration, 3(18), 1. Retrieved from

https://www.westga.edu/~distance/ojdla/

Weiner, J., & Hurtz, G. (2017). A comparative study of online remote proctored versus onsite

proctored high-stakes exams. Journal of Applied Testing Technology, 18(1), 13–20.

http://www.jattjournal.com/

Western Governors University. (2018). Learning at WGU: Different by design, driven by your

needs. Retrieved from https://www.wgu.edu/about/competency-based-education.html#

White, G. (2013). Academic outcomes of developmental community college online and face-to-

face classes: The differences between male and female students. Retrieved from

https://www.capella.edu/

Whitley, B., & Keith-Spiegel, P. (2002). Academic dishonesty: An educator’s guide. Mahwah,

NJ: Lawrence Erlbaum Associates, Inc.

Woods, K. W., & Bliss, K. (2016). Facilitating successful online discussions. Journal of Affective

Teaching, 16(2), 76–92. https://uncw.edu/jet/articles/Vol16_2/index.htm

Xu, D., & Jaggars, S. S. (2011). Online and hybrid course enrollment and performance in

Washington State community and technical colleges. New York: Community College

Research Center, Teachers College, Columbia University.

 103

Xu, D., & Jaggars, S. S. (2014). Performance gaps between online and face-to-face courses:

Differences across types of students and academic subject areas. The Journal of Higher

Education, 85(5), 634–659. Retrieved from https://www.tandfonline.com/loi/uhej20

Yang, Y., Choi, J., & Lee, K. (2018). Theory of planned behavior and different forms of

organizational change behavior. Social Behavior & Personality: An International Journal,

46(10), 1657–1671. https://doi.org/10.2224/sbp.6832

Yates, R. W., & Beaudrie, B. (2009). The impact of online assessment on grades in community

college distance education mathematics courses. American Journal of Distance

Education, 23(2), 62. doi:10.1080/08 923640902850601

Young, J., Birtolo, P., & McElman, R. (2009). Virtual success: Transforming education through

online learning. Learning & Leading with Technology, 36(5), 12–17. Retrieved from

https://searchworks.stanford.edu/view/2999992

Zozus, M. (2017). The data book: Collection and management of research data. Boca Raton, FL:

CRC Press.
 104

Appendix A: Consent Form

Dear CCBC Administration,

As a Mid Atlantic Community College Professor, I am writing to request the opportunity

to collect data for my dissertation in my doctorate program at the American College of Education,

which is based out of Indianapolis, Indiana. I am requesting to be allowed to collect final exam

scores, the percent of successful students, and the percent of students retained in the past three

years (2016 – 2018) who have been in an online institutionalized mathematics course using

MyOpenMath (2018) as the assessment/homework system. I will be comparing the noted scores

and percents with students who had proctored assessments during the semester and those who did

not. I will be comparing the percent of success with traditional classes as well. Therefore, I would

also like to collect the percent of student success (success = C or above) of the above face to face

classes during these same semesters.

The data would be specifically from the following online mathematics classes each

semester (spring, summer, fall, & winter) and grouped by gender and age as well:

Math 081 Pre- Algebra

Math 082 Introductory Algebra

Math 083 Intermediate Algebra

Math 163 College Algebra (Pre-Calc I)

Math 165 Pre- Calculus (Pre-Calc II)

 105

I look forward to your response, and the ability to add to the knowledge base of our

students for continued success.

Sincerely,

Danielle Truszkowski