NURSING RESEARCH

A SEMINAR ON
NON EXPERIMENTAL
RESEARCH DESIGN

Submitted by

Anju joy .k

1
 INDEX

SL NO CONTENT PAGE NO
1 OBJECTIVES

2 INTRODUCTION

3 CONTENT

4 SUMMARY

5 CONCLUSION

6 JOURNEL
ABSTRACT

7 ASSIGNMENT

8 A V AIDS

9 REFERENCE

2
 NON EXPERIMENTAL RESEARCH DESIGNS
CENTRAL OBJECTIVE

At the end of the class students will be able to acquire knowledge regarding non experimental
research designs and will be able to apply this knowledge in their professional area with a
positive attitude

SPECIFIC OBJECTIVES

At the end of the class students will be able to

 define non experimental research

 asses the need of non experimental research
 compare the types
 explain descriptive research design
 describe co relational research design
 discuss about developmental research design
 explain epidemiological research design
 describe survey research design
 discuss about method of data collection
 identify advantages and disadvantages

INTRODUCTION

The research design of a study spells out the basic strategies that researchers adopt to develop
evidence that is accurate and interpretable. The research design incorporates some of the most
important methodologic decisions that researchers make, particularly in quantitative studies.
Thus, it is important to understand design options when embarking on a research project. The
overall plan for addressing a research problem encompasses multiple issues, all of which
have implications for the quality of evidence the study yields. In some studies, nurse
researchers want to test the effects of a specific intervention (e.g., an innovative program to
promote breast self-examination). In such experimental studies, researchers play an active
role by introducing the intervention. In other studies, referred to as non experimental studies,
the researcher observes phenomena as they naturally occur without intervening. There are
numerous specific experimental and non experimental designs from which to choose. Many
research problems cannot be addressed with an experimental or quasi-experimental design.
For example, suppose we were interested in studying the effect of widowhood on health

3
status. Our independent variable is widowhood versus non widowhood. Clearly, we cannot
manipulate widowhood; people lose their spouses by a process that is neither random nor
subject to research control. Thus, we would have to proceed by taking two groups (widows
and nonwidows) as they naturally occur and comparing them in terms of health status

CONTENT

DEFNITION

Non experimental research design is one which the researcher observes the phenomena as
they occur naturally , and no external variables are introduced.

REASONS FOR UNDERTAKING NONEXPERIMENTAL RESEARCH

Most studies involving human subjects, including nursing studies, are nonexperimental. One
reason for using a nonexperimental design is that a vast number of human characteristics are
inherently not subject to experimental manipulation (e.g., blood type, personality, health
beliefs, medical diagnosis); the effects of these characteristics on other phenomena cannot be
studied experimentally.

A second issue is that in nursing research, as in other fields, there are many variables that
could technically be manipulated but could not be manipulated ethically. If manipulating the
independent variable could cause physical or mental harm to subjects, then the variable
should not be controlled experimentally. For example, if we were studying the effect of
prenatal care on infant mortality, it would be unethical to provide such care to one group of
pregnant women while deliberately depriving a second group. We would need to locate a
naturally occurring group of pregnant women who had not received prenatal care. Their birth
outcomes could then be compared with those of women who had received appropriate care.
The problem, however, is that the two groups of women are likely to differ in terms of many
other characteristics, such as age, education, nutrition, and health, any of which individually
or in combination could affect infant mortality, independent of the absence or presence of
prenatal care. This is precisely why experimental designs are so strong in demonstrating
cause-and-effect relationships.

Third, there are many research situations in which it is simply not practical to conduct a true
experiment . Constraints might involve insufficient time, lack of administrative approval,
excessive inconvenience to patients or staff, or lack of adequate funds.

4
Fourth, there are some research questions for which an experimental design is not
appropriate. This is especially true for descriptive studies, which seek to document the
characteristics, prevalence, intensity, or full nature of phenomena. Manipulation is neither
attempted nor considered desirable; the emphasis is on the normal experiences of humans.

Finally, nonexperimental research is usually needed before an experimental study can be

planned. Experimental interventions are developed on the basis of nonexperimental research
documenting the scope of a problem and describing critical relationships between relevant
variables.

TYPES OF NON EXPERIMENTAL RESEARCH DESIGNS

 Co relational research design

 Descriptive research design
 Exploratory research design
 Survey research design

EX POST FACTO/CORRELATIONAL RESEARCH

There are two broad classes of nonexperimental research, the first of which has been called
ex post facto research. The literal translation of the Latin term ex post facto is “from after the
fact.” This means that the study has been conducted after variations in the independent
variable have occurred. Ex post facto research attempts to understand relationships among
phenomena as they naturally occur, without any intervention. Ex post facto research is more
often referred to as correlational research. Basically, a correlation is an interrelationship or
tendency for variation in one variable to be related to variation in another. For example, in
human adults, height and weight are correlated because there is a tendency for taller people to
weigh more than shorter people. Correlational studies often share some structural
characteristics with experimental, quasi experimental, and pre experimental research. If we
use the notation scheme described in the previous section to represent symbolically the
hypothetical study of the effects of widowhood, we find that it bears a strong resemblance to
the nonequivalent control group posttest-only design. The purpose of correlational research,
like experimental research, is to understand relationships among variables. It is, however,
riskier to infer causal relationships in correlational research because of the lack of control
over the independent variable. In experiments, investigators make a prediction that deliberate
variation of X, the independent variable, will result in a change to Y, the dependent variable.

5
For example, they might predict that if a new medication is administered, patient
improvement will result. Experimenters have direct control over the X; the experimental
treatment can be administered to some and withheld from others, and the two groups can be
equalized with respect to everything except the independent variable through randomization.
In correlational research, on the other hand, investigators do not control the independent
variable, which has already occurred. The examination of the independent variable—the
presumed causative factor—is done after the fact. As a result, attempts to draw any cause-
and-effect conclusions are problematic. For example, we might hypothesize that there is a
correlation between smoking and lung cancer, and empirical data would likely corroborate
this expectation: smokers are more likely than nonsmokers to develop lung cancer. The
inference we would like to make is that cigarette smoking causes cancer. This kind of
inference, however, is subject to a fallacy called post hoc ergo propter (“after this, therefore
caused by this”). The fallacy lies in the assumption that one thing has caused another merely
because it occurred before the other. To illustrate why a cause-and-effect conclusion might
not be warranted, let us assume (strictly for the sake of an example) that there is a
preponderance of cigarette smokers in urban areas, and people in rural areas are largely
nonsmokers. Let us further assume that lung cancer is actually caused by poor environmental
conditions in cities. Therefore, we would be incorrect to conclude that cigarette smoking
causes lung cancer, despite the strong relationship shown to exist between the two variables.
This is because there is also a strong relationship between cigarette smoking and the “real”
causative agent, living in a polluted environment. Of course, cigarette smoking/lung cancer
studies in reality have been replicated in so many different places with so many different
groups of people that causal inferences are justified. This hypothetical example illustrates a
famous research dictum: Correlation does not prove causation. The mere existence of a
relationship— even a strong one—between variables is not enough to warrant the conclusion
that one variable caused the other. Although correlational studies are inherently weaker than
experimental studies in elucidating cause-and-effect relationships, different designs offer
different degrees of supportive evidence

types

 retrospective designs
 prospective designs

6
Retrospective Designs

Studies with a retrospective design are ones in which a phenomenon existing in the present is
linked to phenomena that occurred in the past, before the study was initiated. That is, the
researcher is interested in a present outcome and attempts to determine antecedent factors that
caused it. Most of the early epidemiologic studies of the link between cigarette smoking and
lung cancer were retrospective. In such a study, the researcher begins with groups of people
with and without lung cancer (the dependent variable). The researcher then looks for
differences between the two groups in antecedent behaviors or conditions. Retrospective
studies are often cross-sectional, with data on both the dependent and independent variables
collected once, simultaneously. Researchers can sometimes strengthen a retrospective design
by taking certain steps. For example, one type of retrospective design, referred to as a case—
control design, involves the comparison of cases (subjects with a certain illness or condition,
such as lung cancer victims) with controls (e.g., people without lung cancer). In conducting a
strong case—control study, researchers find the cases and obtain from them (or about them, if
records are available) information about the history of the presumed cause. Then the
researchers must find controls without the disease or condition who are as similar as possible
to the cases with regard to key extraneous variables (e.g., age, gender) and also obtain
historical information about the presumed cause. If controls are well chosen, the only
difference between them and the cases is exposure to the presumed cause. Researchers
sometimes use matching or other techniques to control for extraneous variables. To the
degree that researchers can demonstrate comparability between cases and controls with
regard to extraneous traits, inferences regarding the presumed cause of the disease are
enhanced.

Prospective Nonexperimental Designs

A nonexperimental study with a prospective design (sometimes called a prospective cohort

design) starts with a presumed cause and then goes forward in time to the presumed effect.
For example, we might want to test the hypothesis that the incidence of rubella during
pregnancy (the independent variable) is related to infant abnormalities (the dependent
variable). To test this hypothesis prospectively, we would begin with a sample of pregnant
women, including some who contracted rubella during their pregnancy and others who did
not. The subsequent occurrence of congenital anomalies would be assessed for all subjects,
and we would examine whether women with rubella were more likely than other women to

7
bear malformed infants. Prospective designs are often longitudinal, but may also be cross-
sectional (from the subjects’ point of view) if reliable information about the independent
variable is available in records or existing data sources.

Prospective studies are more costly than retrospective studies. For one thing, a substantial
follow-up period may be necessary before the dependent variable manifests itself, as is the
case in prospective studies of cigarette smoking and lung cancer. Also, prospective designs
may require large samples, particularly if the dependent variable of interest is rare, as in the
example of malformations associated with maternal rubella. Another issue is that in a good
prospective study, researchers take steps to confirm that all subjects are free from the effect
(e.g., the disease) at the time the independent variable is measured, and this may in some
cases be difficult or expensive to do. For example, in prospective smoking/lung cancer
studies, lung cancer may be present initially but not yet diagnosed. Despite these issues,
prospective studies are considerably stronger than retrospective studies. In particular, any
ambiguity about whether the presumed cause occurred before the effect is resolved in
prospective research if the researcher has confirmed the initial absence of the effect. In
addition, samples are more likely to be representative, and investigators may be in a position
to impose controls to rule out competing explanations for the results.

Some prospective studies are exploratory. That is, the researcher measures a wide range of
possible “causes” at one point in time, and then examines an outcome of interest at a later
point (e.g., length of stay in hospital). Such studies are usually stronger than retrospective
studies if it can be determined that the outcome was not present initially because time
sequences are clear. However, they are not as powerful as prospective studies that involve
specific a priori hypotheses and the comparison of cohorts known to differ on a presumed
cause. Researchers doing exploratory retrospective or prospective studies are sometimes
accused of going on “fishing expeditions” that can lead to erroneous conclusions because of
spurious or idiosyncratic relationships in a particular sample of subjects.

DESCRIPTIVE RESEARCH

The purpose of descriptive studies is to observe, describe, and document aspects of a

situation as it naturally occurs and sometimes to serve as a starting point for hypothesis
generation or theory development

8
Types

 Comparative descriptive studies

 Univariate descriptive study
 Exploratory descriptive study

Comparative descriptive studies

Although researchers often focus on understanding the causes of behaviors, conditions, and
situations, sometimes they can do little more than describe relationships without
comprehending causal pathways. Many research problems are cast in noncausal terms. We
ask, for example, whether men are less likely than women to bond with their newborn infants,
not whether a particular configuration of sex chromosomes caused differences in parental
attachment. Unlike other types of correlational research—such as the cigarette smoking and
lung cancer investigations— the aim of descriptive correlational research is to describe the
relationship among variables rather than to infer cause-and-effect relationships. Descriptive
correlational studies are usually cross-sectional.

Univariate descriptive studies

Some descriptive studies are undertaken to describe the frequency of occurrence of a

behavior or condition rather than to study relationships. For example, an investigator may
wish to describe the health care and nutritional practices of pregnant teenagers. Univariate
descriptive studies are not necessarily focused on only one variable. For example, a
researcher might be interested in women’s experiences during menopause. The study might
describe the frequency of various symptoms, the average age at menopause, the percentage of
women seeking formal health care, and the percentage of women using medications to
alleviate symptoms. There are multiple variables in this study, but the primary purpose is to
describe the status of each and not to relate them to one another.

Two types of descriptive study from the field of epidemiology are especially worth noting.
Prevalence studies are done to determine the prevalence rate of some condition (e.g., a
disease or a behavior, such as smoking) at a particular point in time. Prevalence studies rely
on crosssectional designs in which data are obtained from the population at risk of the
condition. The researcher takes a “snapshot” of the population at risk to determine the extent
to which the condition of interest is present. The formula for a point prevalence rate (pr) is:

9
Number of cases with the condition or disease at a given point in time

×k

Number in the population at riskof being a case

K is the number of people for whom we want to have the rate established (e.g., per 100 or per
1000 population). When data are obtained from a sample (as would usually be the case), the
denominator is the size of the sample, and the numerator is the number of cases with the
condition, as identified in the study.

Incidence studies are used to measure the frequency of developing new cases. Longitudinal
designs are needed to determine incidence because the researcher must first establish who is
at risk of becoming a new case—that is, who is free of the condition at the outset. The
formula for an incidence rate (ir) is:

Number of new cases with the condition or disease over a given time period

×k

Number at risk of becoming a new case (free of the condition at the outset)

If we continued with our previous example, suppose in october, 2001 we found that 80 in a
sample of 500 people were clinically depressed. To determine the 1-year incidence rate, we
would reassess the sample in october, 2002. Suppose that, of the 420 previously deemed not
to be clinically depressed in 2001, 21 were now found to meet the criteria for depression. In
this case, the estimated 1-year incidence rate would be 5 per 100. Prevalence and incidence
rates can be calculated for subgroups of the population (e.g., for men versus women). When
this is done, it is possible to calculate another important descriptive index. Relative risk is an
estimate of risk of “caseness” in one group compared with another. Relative risk is computed
by dividing the rate for one group by the rate for another. Suppose we found that the 1-year
incidence rate for depression was 6 per 100 women and 4 per 100 men. Women’s relative
risk for developing depression over the 1-year period would be 1.5, that is, women would be
estimated to be 1.5 times more likely to develop depression than men. Relative risk is an
important index in determining the contribution of risk factors to a disease or condition (e.g.,
by comparing the relative risk for lung cancer for smokers versus nonsmokers).

10
Exploratory design

 exploratory design is used to identify, explore &describe the existing phenomenon and its
related factors . In other words, it is not only a simple description or the frequency of
occurrence of a phenomenon, but its in-depth exploration and a study of its related factors to
improve further understanding about a less understood phenomenon . For example, an
exploratory study to assess the multi factorial dimensions of falls & home safety measures
for elderly people living in selected communities in the city.

DEVELOPMENTAL RESEARCH DESIGNS

  Developmental research designs are generally used as adjunct research designs with other
research designs such as cross-sectional-descriptive, longitudinal-correlational research
designs. Developmental research design examines the phenomenon with reference to time.

types

 Cross-sectional design
 Longitudinal design

Cross-sectional design

Cross-sectional research design is one in which the researcher collects data at particular point
of time(one period of data collection). These studies are easier & more convenient to
carryout. For example, a researcher is interested in assessing the awareness on swine flu
among people of an area. Here the researcher interacts only once to collect awareness-
related data from respondents.

Longitudinal studies 

Longitudinal research design is used to collect data over an extended time period (long-time
study). Its value is in its ability to demonstrate change over a period of time. For example, a
researcher in interested in the perception of nursing students towards nursing profession from
the beginning of nursing programme to its end. In this example, it is appropriate to use the
longitudinal research design to study this phenomenon. Longitudinal studies are generally
classified into three types:

I. Trend studies
II. Panel studies

11
 III. Follow-up studies

Trend studies

These studies helps to investigate a sample from a general population over a time with
respect to some phenomenon. Trend studies permit researchers to examine pattern& rate of
changes and to make prediction about future direction based on previously identified.

Panel studies

A panel in research is referred to the sample of people involved in a study. In panel studies,
same people are involved& over a period of time they become more informative on the
phenomenon than the subjects in trends studies because the researcher can not only examine
the patterns of change, but also the reasons for change. & The same selected people are
contacted for two or more times to collect further data

Follow-up studies

These are undertaken to determine the subsequent states of subjects with a specific condition
or those who have received a specific intervention.

SURVEY RESEARCH DESIGN

A survey research design is used to collect information from different subjects within a given
population having same characteristics of interest . If survey is conducted on a sample of
population , it is called simple survey , and if the entire population is involved it is called
population survey such as census, and so on.

It provides superficial information on what people do, eat, seek health care and so on, which
is collected through face to face interview , questionnaire , telephonic or electronic
interviews. It provides extensive rather than intensive results . survey is used to obtain
information about prevalence distributions and interrelations of phenomenon in an population
such as political opinion polls , customer survey and health survey . a survey helps to collect
wide range of data from a given population such as actions , attitudes , opinions ,
perceptions , behaviours , awareness , practices and so on .

12
Main features

 Survey research is a process of gathering current required data from the subjects so
that new information can be obtained .the best feature of the survey research is that is
enables the investigators to easily collect current information about whatever it is they
wish to study.
 In survey research , information is collected from a mix of subjects who represent the
total population in the characteristics being studied.
 Survey research is a mode of enquiry that relays heavily up on the validity of verbal
reports. Surveys can be descriptive , exploratory , comparitative , correlational ,
depending up on the nature of phenomenon being studied.
 Survey data can be collected in a number of ways . The most common method is
questioning. The information is obtained directly from the respondents by self
reporting questionnaires ; however face to face interview method may also be used. A
carefully developed questionnaire or interview schedule is essential for data collection
which must be reliable and valid.
 Personal interviews as regarded as the most useful method of collecting survey data
because of the quality of information that can be obtained. An in-depth response is
possible in an interview as relatively few people refuse to express their views on a
given subject in an interview.

Types of survey

Descriptive survey: it is undertaken to describe the frequency of occurance of a phenomenon

rather than to study relationships.

Exploratory survey : it is the survey of the phenomenon which is very less understood and
helps to operationally define the problem and generate hypothesis.

Comparative survey : comparing and contrasting the existence of a certain phenomenon in

two or more groups is done by comparative survey .

Correlational survey : it is the study of relationship between two or more variable in a

natural setting without manipulation or control.

ADVANTAGE OF NONEXPERIMENTALRESEARCH DESIGN

 Non experimental research designs tend to be closest to real-life situation.

13
  Non experimental research designs are most suitable for the nursing research studies.
 Numerous human characteristics are inherently not subject to experimental
manipulation (e.g. blood type, personality, health beliefs, medical diagnosis, etc.)
 There are many variable that could technically be manipulated, but manipulated is
forbidden on ethical grounds.

DISADVANTAGES OF NONEXPERIMENTALRESEARCH DESIGN

 The major disadvantage of non experimental researches is that the results obtained &
the relationship between the dependent & independent variable can never be
absolutely clear & error-free
 Non experimental studies are conducted for comparative purposes using non
randomly selected groups, which may not be homogeneous & tend to be dissimilar in
different traits or characteristics, which may affect the authenticity& generalizability
of the study results.

SUMMARY

Non experimental research design is also known as observational research design in

biomedical research literature , which is the one of the broad catagories of research designs in
which the researcher observes the phenomena as they occur naturally and no external
variables are introduced . It is a research design in which neither the variables are deliberately
manipulated nor is the setting controlled. In non experimental research , researchers collected
data without making changes or introducing treatments. Data obtained are analysed and the
results may lead to the formation of hypothesis that can be tested experimently. Within the
quantitative framework , the observations are represented by numbers that can be statistically
analysed. Data in non experimental research are generally collected through the
questionnaires , interviews , observations , literature reviews , and critical incident technique.

CONCLUSION

Non experimental research design is one which the researcher observes the phenomena as
they occur naturally , and no external variables are introduced. Research design is the blue
print to conduct a research study, which involves the description of research approach ,study
setting , sample size, sampling technique , tools and method of data collection and analysis to
answer specific research questions or test research hypothesis.

14
JOURNEL ABSTRACT

Nonexperimental research: strengths, weaknesses and issues of precision . Thomas e rieo .

European Journal of Training and Development 40(8/9):676-690 · September 2016 with 
DOI: 10.1108/EJTD-07-2015-0058

Purpose: Nonexperimental research, defined as any kind of quantitative or qualitative

research that is not an experiment, is the predominate kind of research design used in the
social sciences. How to unambiguously and correctly present the results of nonexperimental
research, however, remains decidedly unclear and possibly detrimental to applied disciplines
such as human resource development. To clarify issues about the accurate reporting and
generalization of nonexperimental research results, this paper aims to present information
about the relative strength of research designs, followed by the strengths and weaknesses of
nonexperimental research. Further, some possible ways to more precisely report
nonexperimental findings without using causal language are explored. Next, the researcher
takes the position that the results of nonexperimental research can be used cautiously, yet
appropriately, for making practice recommendations. Finally, some closing thoughts about
non experimental research and the appropriate use of causal language are presented.
Design/methodology/approach: A review of the extant social science literature was consulted
to inform this paper. Findings: Non experimental research, when reported accurately, makes
a tremendous contribution because it can be used for conducting research when
experimentation is not feasible or desired. It can be used also to make tentative
recommendations for practice. Originality/value: This article presents useful means to more
accurately report non experimental findings through avoiding causal language. Ways to link
non experimental results to making practice recommendations are explored.

AV AIDS

Black board

Chart

Flash card

Power point presentation

15
REFERENCE

1. Polit D, Beck C. Nursing research.8 th edition. Philadelphia: Wolters Kluwer

Health/Lippincott Williams& Wilkins; 2012.188-198
2. Burns N, Grove KS. Understanding nursing research.4 th edition. Reed elseveir india
pvt ltd. New delhi ;2007.
3. Basavanthappa BT . nursing research and statistics. 3rd edition. Jaypee brothers
medical publishers. New Delhi.
4. Parahoo k . nursing research 2nd edition. palgrave macmillon newyork.
5. Sharma s k . nursing research and statistics . 3 rd edition . Elsevier publications . new
delhi ;2014. 193-209
6. Nonexperimental research: strengths, weaknesses and issues of precision . Thomas e
rieo . European Journal of Training and Development 40(8/9):676-690 • September
2016 with DOI: 10.1108/EJTD-07-2015-0058

16