Quasi experimental design

The prefix *quasi means “resembling”. Thus quasi-experimental research is research that resembles experimental research but is not true experimental research. Although the independent variable is manipulated, participants are not randomly assigned to conditions or orders of conditions (Cook & Campbell, 1979).

WHAT IS QUASI EXPERIMENT ?

As Campbell and Stanley (1966) explain, quasi-experiments arise when researchers lack the control necessary to perform a true experiment.

Quasi-experiments are recommended when true experiments are not feasible.

“The quasi-experimental studies or situations are nothing but the compromise design”, an apt description when applied to much educational research where the random assessment or  random selection of schools and classrooms are quite impracticable.

Kerlinger (1973)

Quasi-experimental research are carried out through quasi-experimental designs. These designs provide control of when and to whom the measurement is applied, but because random assignment to experimental and control treatment has not been applied, the equivalence of the group is not assured .

Best and Kahn(2006:183)

WHAT IS A TRUE EXPERIMENTAL DESIGN ?

A true experiment has one main component – randomly assigned groups. This translates to every participant having an equal chance of being in the experimental group, where they are subject to manipulation, or the control group, where they are not manipulated.

WHAT IS QUASI-EXPERIMENTAL DESIGN ?

quasi-experiment is simply defined as not a true experiment. Since the main component of a true experiment is randomly assigned groups, this means a quasi-experiment does not have randomly assigned groups.

Why are randomly assigned groups so important since they are the only difference between quasi-experimental design and true experimental design ?

When performing an experiment, a researcher is attempting to demonstrate that variable A influences or causes variable B to do something. They want to demonstrate cause and effect. Random assignment helps ensure that there is no pre-existing condition that will influence the variables and mess up the results.

A common example would be something like, ‘Does chemical X1 cause blindness?’ If you accidentally put all of the people wearing glasses in the condition where you spray X1 in someone’s face, then your results are going to be skewed. This is an extreme and overly simplistic example, but it is demonstrating why normally an experimenter wants to randomly assign people into different groups.

Quasi-experiments are most likely to be conducted in field settings in which random assignment is difficult or impossible. They are often conducted to evaluate the effectiveness of a treatment—perhaps a type of psychotherapy or an educational intervention.

CHARACTERISTICS :

  1. It neither employ pre-experimental design nor the true experimental design in its operation. The design employed is quasi-experimental design.

2. These studies resemble true experimental design as there are two groups identified as experimental and control groups but the groups don’t represent equivalent groups as they are not prepared by randomization or any other valid process as used for true experimental design.

3. There is a similarity b/w true and quasi-experimental design when it comes to comparing the scores of two or more groups. However, different groups and conditions are not created by manipulating an independent variable. Instead, groups are defined on the basis of pre-existing participant or environment variables such as male/female, before and after treatment, studying in the government or public sector, working in the public or private sector.

TYPES OF QUASI EXPERIMENTAL DESIGN :

  1. One group pre-test post-test design

2. Non-equivalent group designs:

    a. the post-test only non-equivalent control group design

    b. the pre-test post-test non-equivalent control group design

3. Time-series design:

     a. Interrupted time series design

     b. Time series with non-equivalent control group design

 1. ONE GROUP PRE-TEST POST-TEST DESIGN :

The researcher estimates the treatment effect by comparing the same individuals( or at least the same pool of individuals) at different points of time, before and after the treatment.

If we assume that behavior of a group is measured both before and after treatment, such an “experiment” can be described as follows:

                                O1        X       O2

where O1 refers to the first observation of a group or pretest, X indicates treatment, and O2 refers to the second observation or posttest.

This one-group pretest-posttest design represents a pre-experimental design or, more simply, may be called a bad experiment. Any obtained difference between the pretest and posttest scores could be due to the treatment or to any of several threats to internal validity, including history, maturation, testing, and instrumentation threats (as well as experimenter expectancy effects and novelty effects).

The results of a bad experiment are inconclusive(not leading to a firm conclusion or result) with respect to the effectiveness of treatment.

Examples: new educational techniqueà interactive learning/computer-assisted learning or outcomes learning to be adopted.

2. NON EQUIVALENT GROUP DESIGNS:

In the Non-equivalent(as there is no random allocation of the two groups that’s why called as non-equivalent) control group design, a treatment group and a comparison group are compared using pretest and posttest measures.

If the two groups are similar in their pretest scores prior to treatment but differ in their posttest scores following treatment, researchers can more confidently make a claim about the effect of treatment.

Threats to internal validity due to history, maturation, testing, instrumentation, and regression can be controlled in a nonequivalent control group design. This is also known as the between-group design.

 A. THE POST-TEST ONLY NON EQUIVALENT CONTROL GROUP DESIGN :

•The treatment is received by a treatment group but not by the comparison group or the two groups might receive alternative treatments, the groups are created by self-selection, administrative decisions, or by some other non – random process. •The groups are only observed after the treatment has been administered. Such a design can be represented as:

             O1   X

         ————–

                O2

01 = Treatment group

02 =  Comparison group

X = Treatment

The post-test difference between the group on the outcome variable is used to estimate the size of treatment effect. However, the internal validity threat of selection usually makes the results of post-test only non-equivalent group design un-interpretable in applied research.

By selection, it means the non-random selection is leading to differences between the groups but not the treatment effect.

B. PRE-TEST POST-TEST NON EQUIVALENT GROUP DESIGNS :

are preferred which are represented as  :

            O1   X   O2

           ————

            O1   X   O2 •The dashed line in both the above representations denotes the non-equivalent nature of the group because of their formation by non-random process.

The pre-test allows us to access whether the groups are equivalent on the dependent measure before the treatment is given to the experimental group. In addition, we can access any changes that may occur in either group after treatment by comparing the pre-test measures for each group with their post-test measures. Thus, not only we can compare the performance of two groups on both pre-test and post-test measures, but we can compare performance within each group from pre-test to post-test.

3. TIME SERIES DESIGN

The defining feature of time series design is multiple observations over time. These observations involve the same participant(example: client’s ratings of counselor trustworthiness after every session) or similar participants.

A. INTERRUPTED TIME SERIES DESIGN :

In this multiple observations occur before and after the treatment. Like

O1 O2 O3 O4 O5 O6 X O7 O8 O9 O10 O11 O12, it shows an equal number of observations on the group ( which is not a mandatory requirement) before and after the treatment is done.

Example: accident rates pre and post safety training ,  error rates pre and post quality training.

B. TIME SERIES WITH NON-EQUIVALENT CONTROL GROUP DESIGN :

•To improve the validity of the time-series design, non-equivalent groups should be used to compare groups over a period of time before and after treatment began. • •An extended series of data collection points are used with both the treatment group and the non-equivalent control group. 

  O1 O2 O3 O4 O5  X  O6 O7 O8 O9 O10

  ———————————————–

  O1 O2 O3 O4 O5  X O6 O7 O8 O9 O10

Simply first, in both control and treatment groups are observed pre and post-treatment initiation, and both of them are compared with each other.

ADVANTAGES :

  1. Quasi-experimental design are more frequently used because they’re more practical and feasible to conduct research.

2. This design is preferred where the sample size is small, and where randomization and availability of a control group are not possible.

3. It’s more suitable for a real natural world setting than true experimental design.

4. This design allows the researcher to evaluate the impact of quasi-independent variables under naturally occurring situations.

5. In some cases hypotheses are practically answered through this design.

DISADVANTAGES :

  1. In this design there is no control over extraneous variables influencing the dependent variable.

2. The absence of a control group and control over the research setting makes the result of this design less reliable and weak for the establishment of a causal relationship between independent and dependent variables.

Share with others

Leave a Reply

Your email address will not be published. Required fields are marked *