NAME - JYOTSANA BAWNE

Quantitative PROGRAMME – M.A APPLIED PSYCHOLOGY

Research SEMESTER - I

YEAR – 1st YEAR

Methods SUBMITTED TO - GOPAL SIR

ASSIGNMENT
 QRM ASSIGNMENT

Q. What is Quasi-experimental research design? Establish its relevance citing appropriate

examples from the empirical evidences.

QUASI EXPERIMENTAL RESEARCH DESIGN

According to Mcburney & White (2007) ‘quasi experiment is a research procedure in which the
scientist must select subjects for different conditions from pre-existing groups’.

According to Broota (1989) “All such experimental situations in which the experimenter does
not have full control over the assignment of experimental units randomly to the treatment
conditions or the treatment cannot be manipulated are called quasi experimental design.”

The term quasi means ‘resemblance’. Thus, quasi experimental design is one that resembles an
experiment but lacks random assignment factor. A quasi-experiment is a prospective or
retrospective study in which units are assigned to treatment conditions in a nonrandom fashion,
such as by administrative decision, self-selection, legislative mandate, or some other
nonrandom process. Quasi experimental designs are sometimes called ex-post facto design or
after the fact experiment, because the experiment is conducted after the groups have been
formed.

DIFFERENCE BETWEEN QUASI EXPERIMENTAL DESIGN AND TRUE

EXPERIMENTAL DESIGN
Experimental design Quasi experimental design
● experimenter has complete control ● experimenter does not have control
over the experiment over the assignment of subject to
● experimenter manipulate variables condition
● role of confounding factors mostly ● manipulation of variable is not possible
controlled ● leaves open the possibility that other
differences exist between the
experimental and control conditions
 ● features usually create less compelling
● support for counterfactual inferences
● researcher has to enumerate alternative
explanations

WHY STUDY QUASI‑EXPERIMENTS?

Firstly, randomized experiments cannot always be implemented, much less implemented with
integrity (West, Cham, & Liu, 2014). Implementing a randomized experiment would often be
unethical in settings such as hospital (HIV patients), people affected with war, natural
calamities etc.
Field settings often do not permit the same degree of control as in the laboratory; as a result,
randomized experiments can become degraded when implemented in the field. Under some
conditions, quasi-experiments can be superior to such corrupted randomized experiments.

THE LOGIC OF QUASI-EXPERIMENTATION IN BRIEF

A causal inference from any quasi-experiment must meet the basic requirements for all causal
relationships:
1) that cause precede effect
2) that cause co vary with effect, and
3) that alternative explanations for the causal relationship are implausible.

Both randomized and quasi-experiments manipulate the treatment to force it to occur before the
effect. Assessing co-variation between cause and effect is easily accomplished in all
experiments, usually during statistical analysis. To meet the third requirement,
quasi-experiments (as it does not use random assignment) rely on other principles to show that
alternative explanations are implausible. Some are:
a) Identification and study of plausible threats to internal validity. Once studied after being
identified, it is likely that they explain treatment-outcome co-variation (Reichardt,
2000).
 b) Primacy of control by design. By adding design elements quasi-experimentation aims
either to prevent the confounding of a threat to validity with treatment effects or to
provide evidence about the plausibility of those threats. The usual alternative to design
controls is statistical controls that attempt to remove confounds from effect estimates
using statistical adjustments after the study is done
c) Coherent pattern matching. It involves a complex prediction about a given causal
hypothesis that few alternative explanations can match.

TYPES OF QUASI EXPERIMENTAL DESIGN

One-Group Posttest Only Designs
In a one-group posttest-only design, observations are collected after a treatment is implemented,
but no empirical comparison condition is implemented to provide a counterfactual contrast.
Because of this, it has been labeled a pre-experimental design (Campbell & Stanley, 1966). It is
a non-empirical assessment. It can be diagramed schematically as
X O
X = treatment
O = an observation
And time flows from left to right.

In this design, a treatment is introduced to a single participant or a single group of participants,

an observation is subsequently made, and that observation (O) is taken as the estimate of the
treatment effect. That is, it is assumed (rather than assessed empirically) that, in the absence of
a treatment effect, the value of the observation (O) would have been zero or some other small
number, with any deviation from zero or that small number being taken as evidence of a
treatment effect.
When a nationwide program to vaccinate people against the swine flu was begun in October
1976, the first groups of people given the vaccine were the elderly. The program administered
shots to 24,000 elderly people during the first week. Eight states stopped the program because
three of those 24,000 died (Freedman, Pisani, & Purves, 1978). They presumably attributed the
three deaths to the vaccines. They would have had to compare the death rates of those who
received the vaccine to the death rates of those who had not received the vaccine.
 Strengths Weaknesses
● Mostly used for qualitative case studies ● does not provide an empirical
(novel education program) counterfactual comparison
● More interpretable under pattern ● highly dependent on assumption
matching ● didn’t identify what would happen
● Identify possible effects of intervention without intervention

For valid descriptive causal inferences to result, the effect must be large enough to stand out
clearly, and either the possible alternative causes must be known and be clearly implausible or
there should be no known alternatives that could operate in the study context (Campbell, 1975).
The one-group design without pretests can be more interpretable under theory linked conditions
variously called pattern matching (Campbell, 1966a; Trochim, 1985) or coherence (Rosenbaum,
1995a). Pathologists use this detective-like approach to identify possible causes of the death by
matching that pattern of data to the descriptions in the scientific literature that differentiate one
disease from another. Epidemiologists do something similar.

Pretest–Posttest Designs
In a basic pretest–posttest design, a pretest measure is observed, the treatment is introduced,
and a posttest measure is observed. Such a design can be diagrammed schematically as
O1 X O2

O1 = pretest observation
X = treatment
O2 = posttest observation
and time moves from left to right

The observations can be made on a single study unit or on multiple study units. For the
treatment effect estimate to make sense, the pretest and posttest must be measured using the
same or parallel instruments. Politicians often use the pretest–posttest design to boast of their
successes, changes made before and after coming to power because of their welfare policies
(Redmond, 2016).

Strengths Weaknesses
● provides an empirical counterfactual ● might not well reveal what would have
comparison happened if the treatment had not been
● often easy to implement implemented
● results are easy for even laypersons to ● estimate of the treatment effect may be
understand biased (plausibility of threats to
internal validity)

The basic pretest–posttest design can be usefully supplemented by adding another pretest
observation. This allows the pattern of growth to be better modeled over time and help rule out
threats to internal validity. Other design variations include using cohorts for the pretest and
posttest samples. Another variation in the basic pretest–posttest design entails the addition of a
nonequivalent dependent variable (Coryn & Hobson, 2011; Shadish et al., 2002). In some cases,
threats to internal validity can be ruled out by collecting auxiliary information.

Non-Equivalent Group Design

In this design both a control group and an experimental group is compared. However the groups
are chosen and assigned out of convenience rather than through randomization. Unlike in
between-groups randomized experiments, in nonequivalent group designs, participants are
assigned to (or choose) treatment conditions non-randomly. Following the administration of the
different treatments in a nonequivalent group design, the participants are assessed on an
outcome (i.e., post treatment) measure. The difference in outcomes between the nonequivalent
groups is then used to estimate the relative effectiveness of the different treatments.
Two basic non-equivalent group designs -
The simplest nonequivalent group design is the posttest-only non-equivalent group design
which is diagrammed as:
NR: X O
NR: O

X = treatment
O = posttest observation
NR and the dashed line = two groups of participants are assigned to treatment conditions
non-randomly.

A slightly more elaborate nonequivalent group design adds a pretest and is therefore called the
pretest–posttest Non equivalent group design. Such a design is diagrammed as:below

NR: O1 X O2
NR: O1 O2

In this design, each of the two groups of participants is assessed on a pretest measure. The
treatment group then receives the treatment while the comparison group receives either no
treatment or an alternative treatment. Then both groups are assessed on the outcome
measure—the posttest.
These designs are widely used in social sciences. For example, Pischke (2007) conducted a
study to assess the effects of the length of the school year on academic performance and
subsequent earnings; Hong and Raudenbush (2005) to assess the effects of retention in grade
level on learning; Rubin (2000) to study of the effects of smoking on health care costs.

Strengths Weaknesses
● One of the most commonly ● Problem with comparing results
implemented research designs in the between experimental and control
social sciences (May, 2012) groups (non random assignment;
● High external validity non-equivalent group)
 ● Easy to implement ● selection differences posed a threat to
internal validity
● produce less credible results

The Separate Pretest - Posttest Sample Design

The basic idea in this design is that the people we use for the pretest are not the same as the
people we use for posttest. The design may be diagrammed as shown below:

NR: O1 | X O2
NR: O1 | O2
the vertical lines indicate that the pretest and posttest observations are collected on separate
samples of participants.

The Double Pre-Test Design

The double pre-test design includes two measures prior to the program. The design may be
diagrammed as:
O1 O2 X O3
This is a very strong quasi experimental design with respect to internal validity. Although the
pre-test helps to assess the degree of pre-program similarity, it does not tell us if the groups are
changing at similar rates prior to the program. Consequently if the program and comparison
group are maturing at different rates we can detect this as a change from pretest 1 to pretest 2.
Therefore this design explicitly controls for selection maturation threats.

The Switching Replications Design

The Switching Replications quasi-experimental design is also very strong with respect to
internal validity. And, because it allows for two independent implementations of the program, it
may enhance external validity or generalizability. The design has two groups and three phases
of measurement. In the first phase of the design, both group are pretested, one is given the
program and both are post-tested. In the second phase of the design, the original comparison
group is given the program while the original program group serves as the “control”.
Interrupted Time-Series Designs
These are the research designs that allow the same group to be compared over time by
considering the trend of the data before and after experimental manipulation. (Mcburney &
White, 2007). In this design pre-testing and post-testing of one group of subject is done at
different intervals possibly in order to determine the long term effect of treatment. This design
can be diagrammed as below:
O1 O2 O3 O4 X O5 O6 O7 O8
The above diagram shows that a series of pre-tests are given to a group. Then treatment (X) is
given and a series of post-tests are given to the same subject. The basic ITS design adds
multiple observations both before and after the treatment implementation, where both the
pretest and posttest observations are strung out over time and all the observations are collected
on the same variable. The data in an ITS design could be collected on a single person or
separately on multiple people. The length of time (i.e., the lag) between observations can be
anywhere from fractions of a second to years. Researchers can generally be most confident of a
treatment effect estimate at the precise point in time when the treatment is introduced and less
confident of an estimate farther removed from the point in time when the treatment is
introduced.

Fig. Different patterns of treatment effect over time. The dashed line indicates when the
treatment was implemented. The solid lines show the trend in observations both before and
after the intervention
 Strengths Weaknesses
● uses a baseline of observations to ● likely to produce less credible results
provide its counterfactual comparison than other designs
● can be implemented with very few ● time consuming
participants or units
● It does not require the treatment to be
withheld from any participants in the
study.
● allow researchers to assess the
temporal pattern of effects more easily
● Power and precision in a basic ITS
design depend mostly on the number
of observations and the heterogeneity
of observations
● tends to produce more credible results
(as it rules out threats to internal
validity)

Multiple Time Series Design

In this design we have two groups, one group receives treatment and the other group does not
receive the treatment and this group acts as the control group. This design can be presented as
given in the diagram below:
Pre response measure Treatment Post response measure
G1 O1 O2 O3 O4 O5 X O6 O7 O8 O9 O10
G2 O1 O2 O3 O4 O5 O6 O7 O8 O9 O10

It is usually a complex setting with many events and trends that might affect the behaviour in
question. The addition of a comparison group can be useful in clarifying the relationship
between the treatment and any change in the series of behavioural measures being used.
Repeated Treatment Design
Repeated treatment design is one in which a treatment is withdrawn and then presented the
second time (McBurney and White, 2007). In this design the treatment is presented more than
once. The subject’s response is measured before and after the introduction of a treatment, then
the treatment is withdrawn and the whole process is began again. The design is shown in
following table-
Pretest1 Treatment Posttest1 Withdraw Treatment Pretest1 Posttest2

Table: A Repeated treatment designdesign

A design can have both a removed treatment and, subsequently, a repeated treatment. Such a
design is often called just a repeated treatment design and would look like the following:
O1 O2 O3 O4 O5 X O6 O7 O8 O9 O10 X O11 O12 O13 O14 O15 X O16 O17 O18 O19 O20
In words, the treatment is introduced, then removed (X), and then reintroduced with
observations collected before and after each of the three interventions. If the effect of the
treatment is transient, the treatment could simply be repeated without having to be explicitly
removed. Such a design would be diagrammed as –

O1 O2 O3 O4 O5 X O6 O7 O8 O9 O10 X O11 O12 O13 O14 O15

An example would be a medication where the effects wear off but the underlying condition that
the medication treats remains over time (such as pain after surgery).

Fig. Types of Quasi experimental design

ADVANTAGES AND DISADVANTAGES OF QUASI EXPERIMENTAL DESIGN

Advantages

● In social science, where pre selection and randomization of groups is often difficult,
they can be very useful in generating results for general trends.
● Quasi-experimental design is often integrated with individual case studies and allows
some sort of statistical analysis to take place.
● In addition, they do reduce the time and resources needed for experimentation.
● They are typically easier to set up than true experimental designs.
● Utilizing quasi-experimental designs minimizes threats to external validity as natural
environments do not suffer the same problems of artificially as compared to a
well-controlled laboratory setting. Findings in one may be applied to other subjects and
settings, allowing for some generalizations to be made about population.
● This experimentation method is efficient in longitudinal research that involves longer
time periods.

Disadvantages

● Without proper randomization, statistical tests can be meaningless.

● The lack of random assignment in the quasi experimental design method may pose
many challenges for the investigator in terms of internal validity. This deficiency in
randomization makes it harder to rule out confounding variables and introduces new
threats to internal validity.
● Because randomization is absent, conclusions of causal relationships are difficult to
determine due to a variety of extraneous and confounding variables that exist in a social
environment.
● Moreover, even if these threats to internal validity are assessed, causation still cannot be
fully established because the experimenter does not have total control over extraneous
variables.
RELEVANCE –

As described earlier, it is useful in generating results for general trends, especially in social
science field. They are often efficient in longitudinal research. For example, addressing
economic challenges compounded for female workers, who often face family-work tensions
due to obligations such as unpaid caring work (Ahl & Nelson, 2015), which are more salient in
poverty contexts in emerging economies (Goodman & Kaplan, 2019; Venkatesh et al., 2017).
The results from this 7-year field quasi-experiment in 20 rural villages in India provided
substantial support that ties to family and community positively, and ties to men in power
negatively, related to information and communication technology (ICT) use, entrepreneurial
activity, and entrepreneurial profit. The results could have implications addressing grand
challenges of empowering women in less developed countries.
In another study, Hazra et al. (2019) demonstrated that health behaviour change integration
intervention through SHGs not only helps improve maternal and newborn practices, but also
reduces disparities between most and least marginalized populations for such practices. This
study looks into the disparities that exist across sub-populations in India. Also, highlighting the
fact that microfinance-based community organizations can be effectively used to create an
enabling environment for appropriate health information to affect health practices and access to
health services, and thereby reduce health inequalities in India, as well as globally. This also
helps in forming pro welfare policies by the respective governments or increasing the
effectiveness of the existing one.

In the context of techno-driven world, with social media and emerging AI technologies
effecting our social context as well as behavior, quasi experimental design play an important
role in establishing a causal relationship among these. For example, employing play therapy as
an effective method and without adverse effects is useful in clinical centers of reducing
aggression and improving spelling disorder (Sarpoulaki & Kolahi, 2016). This helps in
formulating new play based learning curriculum. Additionally, this also promotes the concept of
inclusive education for students with Special Learning Disabilities.
Mohanan & Maskelo (2009) showed evidence of causal effects of a physical health shock on
mental health. This quasi experiment also informs the problem faced by policy makers trying to
allocate resources to physical and mental health issues. Future research efforts targeted at
understanding such spillover effects will help recognize the complementarities in investing in
both mental and physical health and also inform treatment of patients who, having one
conditions, could be at risk of developing the other.
Dixit, Bhardwaj & Aniyan (2020) observed that Self Instructional Module on
Non-Communicable Diseases (NCD) was effective teaching material for Nursing Officers
working in Government Hospitals. It also suggested that the most of Nursing Officers had
average to good knowledge, but due to over burden, patients and lack of time to study, not able
to retain their knowledge. This implies the role of organizations, public and private hospital and
institutions to come up with some effective policies and incentives formulations to ensure
availability of quality services through public health institutions.

Gope et al. (2019) conducted a study to highlights the effects of participatory learning
and action with women’s groups, counseling through home visits and crèches on under nutrition
among children less than three years in eastern India, addressing the high burden of child under
nutrition. These interventions could be scaled up through government plans to strengthen home
visits and community mobilization with Accredited Social Health Activists (ASHA workers),
and through efforts to promote crèches. Another quasi experimental study was employed by
Gauba & Singh (2021) to assess the Effectiveness of a Structured Training Program on
Newborn Care Based on ASHA Module 7 ‑ “Skills That Saves Lives” and focuses on
improving such practices.

In the social context of educational institution children’s mental health problems are associated
with a low self-esteem. A quasi-experimental controlled study of a school based mental health
program can improve the self-esteem of primary school children (Iwahori, Oshiyama &
Matsuzak, 2022). Such program may be helpful as a primarily educational program option to
address mental health problems in primary school children.
Fig. Some areas of quasi experimental design
CONCLUSION

They have the important advantages that typically generate results that are of higher external
validity than experimental results because they take place in ‘‘real world’’ settings rather than in
the artificial context of experiments. Quasi-experiments are also well suited to establish causal
effects on long-term health outcomes, as well as on non health outcomes of a health
intervention, such as social and economic consequences. Quasi-experimental results should
thus be considered and integrated in systematic reviews, meta-analyses, and other evidence
syntheses of causal effects of health care practice, programs, and policy (Tugwell et al. 2007).
Developing and enhancing systematic frameworks for assessing the quality of evidence in
education that includes these new techniques must be a priority for education research
(Gopalan, Rosinger & Ahn, 2020).
REFERENCES

1. Campbell, D. T., Stanley, J. C., & Gage, N. L. (1963). Experimental and

quasi-experimental designs for research. Houghton, Mifflin and Company.

2. Campbell, D. T., Cook, T. D., & Shadish, W. R., Jr. (2001). Experimental and
quasi-experimental designs for generalized causal inference. Houghton Mifflin.

3. B Sarpoulaki, P Kolahi (2016), Role of Play Therapy on Aggression and Learning

Disabilities in Students: A Quasi-Experimental Design. International Journal of Indian
Psychology, 3 (4), DIP: 18.01.021/20160304,DOI: 10.25215/0304.021

4. Dixit S, Bhardwaj UD, Aniyan AT. A Quasi-Experimental Study to Evaluate the

Effectiveness of Self-Instructional Module (SIM) regarding Non-Communicable
Diseases (NCD) among Nursing Officers in Terms of Knowledge. Int J Nurs Midwif Res
2019; 6(4): 8-12.

5. Gauba A, Singh M. A quasi-experimental study to assess the effectiveness of a

structured training program on newborn care based on asha module 7 - “Skills that saves
lives” in terms of reported practice among ASHA workers in a selected community of
Delhi. Indian J Community Med 2021;46:277-80.

6. Gopalan, Maithreyi & Rosinger, Kelly & Ahn, Jee. (2020). Use of Quasi-Experimental
Research Designs in Education Research: Growth, Promise, and Challenges. Review of
Research in Education. 44. 218-243. 10.3102/0091732X20903302.

7. Gope, R. K., Tripathy, P., Prasad, V., Pradhan, H., Sinha, R. K., Panda, R., Chowdhury,
J., Murugan, G., Roy, S., De, M., Ghosh, S. K., Sarbani Roy, S., & Prost, A. (2019).
Effects of participatory learning and action with women's groups, counselling through
home visits and crèches on undernutrition among children under three years in eastern
 India: a quasi-experimental study. BMC public health, 19(1), 962.
https://doi.org/10.1186/s12889-019-7274-3

8. Hazra, A., Atmavilas, Y., Hay, K., Saggurti, N., Verma, R. K., Ahmad, J., Kumar, S.,
Mohanan, P. S., Mavalankar, D., & Irani, L. (2019). Effects of health behaviour change
intervention through women's self-help groups on maternal and newborn health
practices and related inequalities in rural india: A quasi-experimental study.
EClinicalMedicine, 18, 100198. https://doi.org/10.1016/j.eclinm.2019.10.011

9. Iwahori, M., Oshiyama, C. & Matsuzaki, H. A quasi-experimental controlled study of a

school-based mental health programme to improve the self-esteem of primary school
children. Humanit Soc Sci Commun 9, 148 (2022).
https://doi.org/10.1057/s41599-022-01156-x

10. Lin C-L, Huang L-C, Chang Y-T, Chen R-Y and Yang S-H (2021) Under COVID-19
Pandemic: A Quasi-Experimental Trial of Observation on Diabetes Patients’ Health
Behavior Affected by the Pandemic From a Coaching Intervention Program. Front.
Public Health 9:580032. doi: 10.3389/fpubh.2021.580032

11. Mohanan, M., & Maselko, J. (2010). Quasi-experimental evidence on the causal effects
of physical health on mental health. International journal of epidemiology, 39(2),
487–493. https://doi.org/10.1093/ije/dyp331

12. Reichardt, C. S., & Mark, M. M. (1998). Quasi-experimentation. In L. Bickman & D. J.

Rog (Eds.), Handbook of applied social research methods (pp. 193–228). Sage
Publications, Inc.

13. Venkatesh, Viswanath & Shaw, Jason & Sykes, Tracy Ann & Macharia, Mary & Fosso
Wamba, Samuel. (2017). Networks, Technology, and Entrepreneurship: A Field
Quasi-experiment among Women in Rural India. The Academy of Management Journal.
60. 1709-1940. 10.5465/amj.2015.0849.