INTRODUCTION TO

QUANTITATIVE RESEARCH

ADONIS P. DAVID, PhD

Philippine Normal University
Content

 Quantitative Research Designs

 Quantitative Research
Instrumentation
 Quantitative Data Analysis
 Current Trends in Qualitative
Research

2
Types of Research Designs
Research Designs

Quantitative Qualitative Analytical Study Mixed Method

Case Study Concept Analysis

Non-Experimental Experimental
Phenomenaology Historical Analysis

Descriptive True Ethnography

Comparative Quasi Grounded Theory

Correlational Single Subject

Causal Comparative
Quantitative research
methodologies

 require the collection and analysis of

numerical data from measurements
of research subjects or participants.
They are often used to test theories,
hypotheses, assumptions or ideas to
determine their validity or accuracy.
Quantitative Research
 Non-experimental- Survey of some part
of a population
 A “sample” via questionnaire (if people are
involved)
 Results enables researcher to generalize the
findings from a sample of responses.
 Experimental- TESTS cause and effect
relationships between groups.
 Researcher manipulates one or more
independent variables
 Determines if manipulations cause an
outcome
Quantitative Research Designs
 Experimental

(a) Examines and establishes cause

and effect relationships
between variables using
experimental methods

(b) Involves manipulation of

independent variables and
control of extraneous variables
Quantitative Research Designs
 Non-experimental

(a) Investigation of current status of

variables and their relation with
other variables

(b) Does not involve manipulation of

variables
 Classifications of Experimental
Designs
Classification Purpose Typical Role of Sample Research
Objectives Hypothesis Question

True Experimental Investigate cause Determine the Hypothesis testing What is the effect
and effect effects, examine or confirmation of a behavioural
relationships with the differences, modification
samples randomly compare the program on the
assigned to groups results, or analyze participants from
the influence the experimental
group?
Quasi- Investigate cause Determine the Hypothesis testing What is the effect
experimental and effect effects, examine or confirmation of a behavioural
relationships the differences, modification
without random compare the program on
assignment of results, or analyze students of Class
samples to groups the influence A?
Non-Experimental Designs
Classification Purpose Typical Role of Sample Research
Objectives Hypothesis Question

Descriptive Describe the describe, No hypothesis To what extent

current status determine, are high
and patterns analyze, school
or explain, teachers using
characteristics define, concept
of a research illustrate notes?
variable
Correlational Determine the correlate, May or may What are the
extent to interrelate, not have predictors of
which two or associate, hypothesis teachers’
more variables relate, burnout?
are statistically determine
related effect
 Non-Experimental Designs
Classification Purpose Typical Objectives Role of Hypothesis Sample Research
Question

Causal-Comparative Examine causal Compare or contrast Typically have What is the effect of
relationships among , determine hypothesis family structure on
variables that cannot similarities, find out students’ test
be actively differences, performance?
manipulated or determine effect Is there a significant
controlled by the difference in the
researcher ; academic motivation
Confirms if two or of students who are
more groups have exposed to domestic
similar or different violence and those
patterns or who are not?
characteristics when
compared using a
set of variables as
basis
 Non-Experimental Designs
Research Objective Time Dimension

Descriptive Retrospective
Predictive Cross-sectional
Explanatory Longitudinal
Examples:
Research Title Design

Psychological predictors of school-based violence: ???

Implications for school counselors

Testing a model on why students’ misbehaved in social media ???

Steps in the Process of Quantitative
Data Collection
 Determining participants to study
 Obtaining permissions needed
 Considering what types of information to
collect
 Locating and selecting instruments
 Administering the data collection
Information to Collect:
Types of Data Measures

 An instrument is a tool for measuring, observing, or

documenting quantitative data.
 Types of instruments
 Performance measures (e.g., test performance)
 Attitudinal measures (measures feelings toward
educational topics)
 Behavioral measures (observations of behavior)
 Factual measures (documents, records)
Characteristics of Good
Measures
 Is the measure relevant?
 Is the measure credible?
 Is the measure valid?
 Is the measure reliable?

15
Types of Measures
 Cognitive Measures

 Non-cognitive Measures
 Questionnaires
 A structured way of asking questions and eliciting
answers or responses, typically through a paper-
and-pencil approach.
 Questionnaires can be designed and used to
collect vast quantities of data from a variety of
respondents. They have a number of benefits over
other forms of data collection: they are usually
inexpensive to administer; very little training is
needed to develop them; and they can be easily
and quickly analysed once completed.
Type of Questions

5/27/2019 18
S

5/27/2019 19
S

5/27/2019 20
 Scaling Technique
 Some questions require the respondent to indicate
answers according to a predefined list or scale,
usually ranging from a very positive answer to a
very negative answer.

 There are a number of ways to scale responses to

questions. One of the most popular approaches is
the Likert scale (measure of position in an opinion)

5/27/2019 21
Why Use Questionnaires?
 Questionnaires can facilitate the
collection of vast amounts of data
with minimal effort.

 Well-designed questionnaires can

allow relationships between data to
be identified. They are particularly
useful for showing relationships with
data that are easily quantifiable.
5/27/2019 22
 Questionnaires protect respondent
anonymity: they can be distributed
and returned confidentially and
without the respondent ever being
identified.
 As research instruments,
questionnaires can be used time and
time again to measure differences
between groups of people. They are
reliable data gathering tools..
5/27/2019 23
Sample Research Questionnaire

24
Sample Research Questionnaire

25
Before moving to test
development…
 Adopt?

 Adapt?

 Adapt and Translate?

 Develop?

26
Adopt
 Adopt and use an existing test or
questionnaire with little or no
modification/revision.

27
Adapt
 Adapt and use an existing test or
questionnaire with minor
modifications/revisions usually in
terms: (a) item content or (b)
response option.

28
Adapt and Translate
 Adapt by translating an existing test
or questionnaire written in English or
other foreign language to Filipino or
some other Philippine language or
dialect.

29
Develop
 Designing and constructing a test or
questionnaire

30
Perceived Academic Control Scale

5/27/2019 31
 Quantitative Data Analysis
(Statistics)

Parametric Non-Parametric
•Enables researchers to make •Difficult to make assumptions
assumptions about the population about the population
•Large sample size is requires •Large sample size is not a
(N>30) requirement
•Used for interval and ratio scales •Used for nominal and ordinal
scales
 Statistics and number of groups
Design Parametric Non-Parametric
One sample z-test One-way chi-square
-the mean of one sample is t-test Kolmogorov smirnov
compared with a standard
One sample repeated t-test for 2 McNemar change test
measures (dependent groups) dependent samples Wilcoxon signed ranks
-One sample is studied test
measured twice (2 set of data)
- e. g. pre and post test design
Two independent groups t-test for 2 Two-way chi-square
-studying two distinct independent Mann Whitney U test
samples/groups samples
Comparing multiple groups Analysis of Variance Kruskal wallis test
(independent or dependent (ANOVA)
groups) 1 IV, 1 DV: one way
ANOVA
2 IV, 1 DV: two way
ANOVA
1 more IV, 2 or
more DV: MANOVA
Statistical Analysis
Statistical Approach Focus Sample Research Question

Measures of Central Tendency Description of the variables What is the level of X?

Frequeny and Percentage How much percentage refers to

X?

How many are above average

Standard scores
in terms of X?
Correlation Analysis Relation between two Is there a relation between X
variables and Y?

Regression Analysis Relation of two or more Are X1, X2, and X3 associated
predictors with an outcome with Y?
variable
Do X1, X2, and X3 significantly
influence Y?
Statistical Analysis
Statistical Approach Focus Sample Research Question

Canonical Correlation Analysis Relation of a set of variables Is there a significant

with another set of variables association between Xn and
Yn?
Discriminant Analysis Relation of 1 or more Does (continuous )X predict
continuous predictors on a (categorical) Y?
categorical outcome

IV = continuous, DV =
categorical
Mediation Analysis Mediating effect of M on the Does M mediates the
association of X and Y association between X and Y?

Moderation Analysis Moderating or interacting Does M moderate the

effect of M on the association association between X and Y?
of X and Y?
 Statistical Analysis
Statistical Approach Focus Sample Research Question

t-test Significant difference between Is there a significant difference

Group 1 and Group 2 between the two groups?

ANOVA Significant difference between Are there significant

two or more groups differences among the groups?

Factorial ANOVA Significant difference between Are there significant

2 or more grouping variables differences among the S in
terms of the grouping
variables?
Repeated Measures ANOVA Significant difference between Do the conditions have an
two or more conditions effect on Y?
experienced by the same group
of S
Statistical Analysis
Statistical Approach Focus Sample Research Question

Mixed ANOVA Significant difference between Do the conditions and

two or more groups or grouping variables have an
grouping variables and effect on Y?
between 2 or more conditions
ANCOVA Significant difference between Are there significant
two or more groups while differences among the groups
controlling other variables even when controlling the
value of C?
MANOVA Significant difference between Are there significant
groups or grouping variables differences among the groups
on 2 or more outcomes across 2 or more outcomes?
MANCOVA Significant difference between Are there significant
groups or grouping variables differences among the groups
on 2 or more outcomes while across 2 or more outcomes
controlling other variables even when controlling for the
value of C?
CURRENT TRENDS IN QUANTITATIVE
RESEARCH

1. Structural Equation Modelling

2. Person-Centered Analysis
3. Latent Growth Modeling
CURRENT TRENDS IN QUANTITATIVE
RESEARCH

1. Effect size is reported.

2. Power is reported.

3. Instrument validity and reliability are reported based

on data from the current study.

4. Longitudinal data analysis.

5. Combining correlational and experimental methods.

STRUCTURAL EQUATION
MODELLING (SEM)
Goes beyond regression models: Test
several variables and latent constructs
with their underlying manifest variables.

Provide a way to test the specified set

of relationships among observed and
latent variables as a whole and allow
theory testing for causality even when
independent variables are not
experimentally manipulated
STRUCTURAL EQUATION
MODELLING (SEM)

The entire model is tested if the data

fits the specified model.

Takes into account errors of

measurement.
Variables in a Structural Model:

1. Manifest variables: Directly observed or

measured. Boxes are used to denote
manifest variables.

2. Latent variables: Not directly observed;

we learn about them through the
manifest variables that are supposed to
“represent” them. Ovals are used to
denote latent variables.
Path Analysis
Self-efficacy
.48
* E
1
1.0
.28
Deep Approach Metacognition
*

.20
*

Surface
Approach

2=10.03, df=3, 2/df=3.34, GFI (.98), adjusted GFI (.92),

RMSEA= .08
SEM Analysis
DELTA DELTA DELTA DELTA DELTA DELTA
2 3 4 5 6 7

100.4 71.4 57.1 34.9 71.9 88.1

3*
Conditiona 6* 1*
Planni 4*
Monitor 2*
Informatio 0*
Debuggin
l Procedural n g Strategy
Knowledg Knowledg ng ing Managem
e e 6.88 ent
7.07 9.25 * 7.91
* * * 7.24
Declarative 9.03 * Evaluati
Knowledge
* 6.27 on
* Metacogni 25.1
82.5 tion 2* 78.3
7*
DELTA 9*
DELTA
1 8
2.10
* 5.19
Critical
* ZETA
Thinking
1

0.67 0.86 0.74 0.40

* * * *
Infere Recognition Deducti Interpret Evaluatio
nce of on ation n of
Assumption Argumen
7.27 2.06 6.15 5.03 ts 3.57
*
EPSILON *
EPSILON *
EPSILON *
EPSILON *
EPSILON
1 2 3 4 5
PERSON-CENTERED ANALYSIS

Variable-centered analysis
 Multiple regression
 CFA (confirmatory factor analysis)
 SEM (structural equation model)
Limitations of the variable-centered
approach
 Isolates psychological traits on which
individuals reliably differ and then study
their correlations with other related
variables
 Variable-centered approaches fail to
provide information about person-
specific intra-individual organization of
psychological processes and behavior
instead they provide information about
the psychological processes for an
average person in the sample (Asendorpf, 2010)
Advantages of a person-centered
approach
 Encourages researchers to think
about configurations of traits within
individuals; whereas a variable-
centered approach is silent about the
intra-individual structure of
psychological factors

 Investigates properties of a person

vs. properties of the average person
in a sample
PERSON-CENTERED ANALYSIS

 Cluster analysis
 Discriminant analysis
 Latent class/ latent profile
analysis
Cluster analysis
 A class of techniques used to classify
cases into groups based on a defined
set of variables
 Members of each group (cluster) are
relatively similar to each other and
different from those from other
groups
 cluster: A group of relatively
homogeneous cases or observations
LATENT GROWTH MODELING

- A structural equation modeling

approach to the analysis of change.

-We need a developmental model,

which takes into account development
and the differences in development
across individuals.
-- The latent factors in the growth
model are interpreted as common
LATENT GROWTH MODELING

In LGM, repeated measures of a

variable (Y) are treated as indicators of
latent variables, called basis curves,
that represent aspects of change—
typically intercept and linear slope
factors. Values of the time metric (e.g.,
age, day, or wave of measurement) are
built into the factor loading matrix to
reflect the form of the hypothesized
trajectory, or trend over time.
 Linear Growth Curve Models

 We have at least three time points for each individual.

 We fit a straight line for each person:

30
25
Outcome

20
15
10
5
0
0 2 4 6 8 10
Time

 The parameters from these lines describe the person.

52
The Key Parameters

 Slope: the rate of change

 Some people are changing more than others
and so have larger slopes.
 Some people are improving or growing
(positive slopes).
 Some are declining (negative slopes).
 Some are not changing (zero slopes).
 Intercept: where the person starts
 Error: How far the score is from the line.
53
Latent Growth Models (LGM)

 For both the slope and intercept there is a mean

and a variance.
 Mean
 Intercept: Where does the average person
start?
 Slope: What is the average rate of change?
 Variance
 Intercept: How much do individuals differ in
where they start?
 Slope: How much do individuals differ in their
rates of change: “Different slopes for different
folks.”
54
Measurement Over Time
 measures taken over time
 chronological time: 2006, 2007, 2008
 personal time: 5 years old, 6, and 7
 missing data not problematic
 person fails to show up at age 6
 unequal spacing of observations not
problematic
 measures at 2000, 2001, 2002, and
2006
55