QUASI-EXPERIMENTAL RESEARCH: CHARACTERISTICS, METHODOLOGY - SCIENCE - 2024

The i esearch quasiexperimental encompasses those studies performed without there being a random assignment groups. It is usually used to determine social variables and some authors consider it unscientific. This opinion is given by the characteristics of the subjects studied.

The non-randomness in their choice determines that there will be no control over the important variables. Likewise, it makes this type of research much more prone to the appearance of biases. There are a number of alternatives when designing the study.

For example, you can establish historical controls or, although it is not mandatory, create a control group that serves to verify the validity of the results. It is considered that this type of research can be divided into four types: natural experiments, studies with historical controls, post-intervention studies and before / after studies.

The method has a number of advantages and disadvantages. Among the first, the ease and economy of carrying them out stands out, apart from being able to be applied to individual situations.

Among the latter is the already mentioned lack of randomness when choosing the groups and the possible appearance of the so-called placebo effect in some of the participants.

characteristics

The origin of quasi-experimental research was in the educational field. The very characteristics of this sector prevented studies of certain phenomena from being carried out with conventional experiments.

Starting in the 60s of the last century, but especially in recent decades, this type of study has multiplied. Today they are very important in applied research.

Manipulation of the independent variable

As is also the case in experimental research, these studies aim to define how an independent variable acts on the dependent one. In short, it is about establishing and analyzing the causal relationships that occur.

Non-random groups

As noted above, one of the defining characteristics of quasi-experimental research is the non-randomization in the formation of groups.

The researcher resorts to groups already formed by the circumstances that are. For example, they can be the members of a university class or a group of workers who share an office.

This causes that there is no certainty that all subjects present similar characteristics, which can cause the results to be not totally scientific.

For example, when it comes to studying school feeding and related allergies, there may be totally healthy children who can distort the results.

Little control of variables

These models are common in applied research. This means that they will be developed in environments other than laboratories, in natural contexts. In this way, the researcher's control over the variables is much less.

Methodologies

In short, the way in which quasi-experimental investigations are carried out is very simple. The first thing is to choose the group to study, after which the desired variable is assigned. Once this is done, the results are analyzed and conclusions are drawn.

To obtain the desired information, various methodological tools are used. The first is a series of interviews with individuals from the chosen group. Similarly, there are standardized protocols for making the pertinent observations that ensure a more objective result.

Another aspect that is recommended is to do a “pre-test”. This consists of measuring the equivalence between the subjects studied prior to the experiment.

In addition to these general lines, it is important to clearly delimit the type of design that you want to establish, since it will mark the direction of the investigation.

Cross-sectional designs

They serve to compare different groups, focusing the investigation on a specific time point. Thus, it is not used to obtain universal conclusions, but simply to measure a variable at a specific time.

Longitudinal designs

In this case, several measures of the variable will be taken for each individual. These, who are the subjects of the study, can range from a single person to groups that make up a unit, such as a school.

Unlike what happens with the transversal ones, this design aims to study the processes of change in a continuous period of time.

Advantages and disadvantages

Advantage

In many social science studies it is very difficult to select groups that can meet the requirements for purely experimental investigations.

For this reason, quasi-experimentals, although less precise, become a very valuable tool for measuring general trends.

A very classic example is the measurement of the effect of alcohol in adolescents. Obviously, it would not be ethically possible to give the kids a drink and observe the effects experimentally. So what researchers do is ask how much alcohol they have drunk and how it has affected them.

Another advantage is that these designs can be used in individual cases and, later, extrapolated with other similar interviews.

Finally, the characteristic of these studies makes them much cheaper and easier to develop. The resources required and the preparation time is much less than if you wanted to perform a traditional experiment.

Disadvantages

The main disadvantage that experts point out is not bringing the groups together randomly, at random. This causes that the results may not be as exact as would be desired.

Part of the problem is the impossibility of researchers to take into account external factors that can distort the responses of the subjects.

Any pre-existing circumstance or personal trait that does not suit the study may lead to different conclusions. Then, the researcher is left without response to these situations.

On the other hand, many theorists warn that what they call the placebo or Hawthorne effect can occur. This consists of the possibility that some of the participating subjects change their behavior when they know that they are participating in a study.

It is not that there is external manipulation, but it has been shown that human beings tend to adapt their behavior to general patterns or to what they think is expected of them.

To try to prevent this from altering the results, the researchers have methodological tools to avoid it, although 100% control is impossible.

References

1. Bono Cabré, Roser. Quasi-experimental and longitudinal designs. Recovered from diposit.ub.edu
2. Migallón, Isidro. Quasi Experimental Research: Definition and Designs. Obtained from psychocode.com
3. Jaen University. Quasi-experimental study. Obtained from ujaen.es
4. Trochim, William MK Quasi-Experimental Design. Retrieved from socialresearchmethods.net
5. Statistics Solutions. Quasi-experimental Research Designs. Retrieved from statisticssolutions.com
6. Research Connections. Experiments and Quasi-experiments. Retrieved from researchconnections.org
7. Wikieducator. Quasi-experimental Research. Retrieved from wikieducator.org