18 julio, 2024

Internal validity: how it is achieved, threats, examples

The internal validity It is a fundamental concept for the research methodology, since it determines the degree of reliability of the experiments and the interpretations that are extracted from them.

An experimental research process needs strict control of all influencing factors for it to have internal validity. When we talk about control, we refer to the accurate knowledge of the relationship between the independent and dependent variables and how they affect each other to determine the results.

That is, the control allows us to identify the causality of the changes that occurred in the variables of an experiment.

To achieve this, other variables that do not apply within the hypothesis being tested must be prevented from intervening in the change of the dependent variables; Only then will it be known if the independent variables influence them.

To achieve internal validity, it is necessary to isolate the specific relationships under study between the independent and dependent variables, to prevent the experiment from becoming «contaminated.»

[toc]

How internal validity is achieved

To achieve control —and therefore, the internal validity of an experiment— the first thing to keep in mind is that you must have a minimum of two experimental comparison groups.

If it were experimented with a single group it would be impossible to know if there was no other influencing factor outside the independent variable that was manipulated. For example, to find out if a fertilizer has an effect on the growth of a plant, it is necessary to compare the treated plant with another that has not been.

Furthermore, these comparison groups must be exactly the same in all but the way the independent variables being tested are manipulated.

If control groups are known to be the same in all but the way they were exposed to the independent variables, any changes they undergo in the experiment must be attributed to these variables; that is, it would be known that the dependent variables were caused by the independent ones.

Threats to internal validity

The possible sources of internal invalidation are external explanations to the variables contemplated within the experiment and that threaten the reliability of the conclusions reached by the investigation.

Interruption by external factor

The first great threat is that some of the participants or objects of study suffer some event unrelated to the experience of the others during the tests. This would eliminate the equivalence of the experimental and control groups.

Therefore, the scientist must ensure that each of the objects of study experience exactly the same events.

Instrumentation

Another threat to internal validity is the instability of the outcome measurement instrument.

To prevent this from affecting the experiment, it is necessary to previously check the stability of the instrument, repeating several tests measured with said instrument over an extended period and verifying repeatability patterns without anomalies in the results.

It is also necessary to take into account that the measurement instrument must be the same for each experimental group.

The experimental environment

Apart from the measuring instrument, the experimental environment must also be taken into account. This must be controlled and it must be ensured that all study objects, experimental and control, are under the same conditions.

Human factors

It must be verified that at the time of beginning the experiment all the participants or objects of study present normal data in terms of the measured variables, that they are not undergoing a process that alters the real assessment of the characters studied.

Another possible threat is that the study subjects interrupt the research by abandoning it in the middle of the process. To solve this inconvenience it is necessary to replace the subject with one that is similar.

The human factor in experimental research processes is one of the most unstable. The researcher must try to keep the study subjects motivated by means of compensations so that as far as possible the subjects are the same from the beginning to the end of the investigation.

If the subjects studied are human beings, it is necessary to take care that they do not communicate with each other, since the information that they can share around the different variables that they experience can affect the natural development of the investigation.

Another human factor to take into account (apart from the attitude of the study subjects) is the attitude of the researcher himself. This must at all times seek objectivity, behave in the same way and perform the same procedures with all the subjects and objects of study.

Examples of internal validity

Example 1

Suppose you want to investigate the effect of a television commercial on the consumer’s willingness to purchase the advertised product.

To carry out a valid experiment in this case, there should be at least two groups: one that has seen the commercial and another that has not seen it.

In addition, external variables should be controlled. It could be the case that some of the study subjects have heard about the product from their friends or have previously tried it and, therefore, know first-hand its characteristics and qualities.

These are aspects that would affect the consumer’s perception of the product and have nothing to do with the independent variable studied: exposure to the commercial. For this reason, the ideal would be to choose study subjects who have not been exposed to these variables.

Example 2

Another example can be an investigation on the influence of a pedagogical method in the learning process.

For a study of this type, the equivalence of the study subjects is extremely important, both in the experimental and in the control groups, since variables such as disparity in the intellectual capacity of the participants could be present.

Before carrying out the experiment, the possibility of excessive disparity in the readiness of the subjects for learning must be ruled out; otherwise, the study would lack internal validity.

Reference

«Stability» (sf) in Infas Control. Retrieved on July 11, 2019 from Infas Control: infas.com.ar
“Internal validity” (nd) in Indiana. Retrieved on July 11, 2019 from Indiana: indiana.edu
Baptista, P., Fernández, C. & Hernández Sampieri, R. “Research Methodology” (2014). Mexico City: McGraw-Hill/Interamericana
Cepeda, M. and Quezada, M. “Research Design, Internal Validity, and External Validity” (March 26, 2016) on SlideShare. Retrieved on July 11, 2019 from SlideShare: en.slideshare.net
Cuncic, A. “Understanding Internal and External Validity” (June 20, 2019) in VeryWell. Retrieved on July 11, 2019 from VeryWellMind: verywellmind.com

Deja una respuesta

Tu dirección de correo electrónico no será publicada. Los campos obligatorios están marcados con *