Dr S. Shyam Prasad
Introduction
Recently, or I feel so, there is a spurt of papers written by management students. Most of them are termed as āresearch papersā. When they are guided by able teachers, they turn out to be very interesting and a good read; they get published too. However, many students are lacking in the knowledge of validity. Some of them donāt even possess a basic understanding of the term validity. Hence I thought of writing this blog; it is an introductory article. Though it is mainly intended for the students writing serious research papers, it will come handy to most of us, including me.
Validity
This word is commonly used in research. In my opinion, it is the foundation of any research. A good understanding of the term would make one confident of oneās research and will stand the test. Kelly (1927) introduced the idea of validity when he said that a test is valid if it measures what it claims to measure. In simpler terms, validity can be understood as the truthfulness of the inference or conclusion of the research. Many students when talking about validity in research, say that a measure is valid or a questionnaire is valid or a sample is valid. However, this is technically wrong. āMeasures, samples, and designs donāt have validity ā only propositions can be said to be validā (Trochim, 2003) .
I would like to draw attention of the reader to the point that the concept of validity is applicable in quantitative research and the below definitions and discussion may not apply to the qualitative research. In qualitative research the equivalent terms are precision (Winter, 2000) , credibility, and transferability (Hoepfl, 1997) whose discussion is out of scope of this blog.
Types of Validity
Basically, there are four types of validity. 1. Construct Validity, 2. Internal Validity, 3. Conclusion Validity and 4. External Validity.
To discuss them with an example, let us assume that we are trying to find out if a 6-day orientation programme in ISME increases the chances of making the admittees better students. This is an idea/thought that is still in our mind. We need to operationalise it. The 6-day orientation can be operationalised by conducting 6-day classes and if a student is better can operationalised by considering studentās CGPA. In this case, ā6-day orientation programmeā is known as the cause construct and ābetter studentā is the effect construct.
1. Construct Validity
Construct validity measures how truthfully we are measuring a given construct. In our example, does CGPA really measures the degree of a student being better? Judging construct validity is a huge challenge. One has to consider many aspects to establish the construct validity such as pilot studies, convergent validity (the degree to which it is similar to other similar measures), discriminative validity (the degree to which it adequately differentiates itself), and so on. There is lots of scope for subjectivity and has no real scale to measure it. It is open to debate. It is mostly subjective.
2. Internal Validity
Internal validity refers to the degree to which the independent variable can accurately be said to produce the observed effect. In our case, how accurately can we say that the betterment of the students was due to the orientation programme? Have we considered the confounding variables
[1]? Have we accounted for students IQ, his previous academic performances etc.?
[1]? Have we accounted for students IQ, his previous academic performances etc.?
One way for increasing the internal validity is to have random samples and using control variables that would account for the betterment of students. That is if we have accounted for the confounding variables the internal validity would be high. However, if we fail to do so, the internal validity would be very low.
3. Conclusion Validity
The most important of all the four validities is the conclusion validity. Conclusion validity comes into play whenever the researcher tries to conclude if there is a relationship in the variables or observations. In fact, this was the whole purpose of the research. āConclusion validity is the degree to which conclusions we reach about relationships in our data are reasonableā (Trochim, 2003) .
In our example, let us say based on data, we conclude that thereās a positive relationship between the orientation programme and the students being better. Conclusion validity will tell how reliable our conclusion is.
A word of caution is that we can never have perfect validity; if one recollects, Type 1 and Type 2 errors are part of the hypothesis testing process. Hence, we can conclude that conclusion validity refers to reasonable conclusions based on the data.
4. External Validity
External validity refers to the extent to which the findings of a research can be generalised i.e. the findings can be extended beyond the sample.
In our example, we found a positive relationship between a 6-day orientation programme and the students are better; external validity refers as to how far can we extend this finding to other institutes and colleges. External validity can also be increased by using random sampling.
Besides the above four types of validity, there are numerous others; few of them are 1.Face validity, 2. Content validity, 3. Predictive Validity, 4. Concurrent Validity and 5. Criterion-related validity. The list is not exhaustive; it is beyond the scope of this blog to discuss all them. Hence I have included only the most important and basic types of validity. However, before closing, I thought it would be prudent to discuss the threats to validity.
Threats to validity
Threats to validity refer to different reasons why a conclusion may be wrong. Let us assume that in our example suppose we didnāt find any relationship between the two variables. How could this conclusion be wrong? There are several reasons.
1. It is possible that there isnāt sufficient statistical power to detect a relationship.
2. The sample size was very small.
3. The assumptions of the correlational test are violated
4. Random irrelevancies in the study setting
5. Unreliable measures and
6. Random heterogeneity in the respondents.
Conclusion
To produce a good research paper it is indispensable and well worth understanding the theory of validity. One passing remark is that along with validity, the term reliability is closely associated with it. Reliability refers to how consistently a method measures something. In other words, if one uses the same method repeatedly in different settings and if the same results are got then that method is said to be reliable.
I thank the reviewer and Dr J Meenakumari of their valuable suggestions and inputs.
Disclaimer: The views, opinions, and content on this blog are solely those of the authors. ISME does not take responsibility for the content which are plagiarized or not quoted.
References
1. Glen, S., 2014. Construct Validity: Simple Definition, Statistics Used. [Online]
Available at: https://www.statisticshowto.com/construct-validity/
[Accessed 07 June 2020].
Available at: https://www.statisticshowto.com/construct-validity/
[Accessed 07 June 2020].
2. Hoepfl, M. C., 1997. Choosing qualitative research: A primer for technology education researchers. Journal of Technology Education, 9(1), pp. 47-63.
3. Kelley, T. L., 1927. Interpretation of educational measurements. New York: Macmillan.
4. McLeod, S. A., 2013. What is validity?. [Online]
Available at: https://www.simplypsychology.org/validity.html
[Accessed 07 June 2020].
Available at: https://www.simplypsychology.org/validity.html
[Accessed 07 June 2020].
5. Trochim, W. M. K., 2003. Research Methods. 2 ed. New Delhi, India: biztantra.
6. Winter, G., 2000. A comparative discussion of the notion of validity in qualitative and quantitative research. The Qualitative Report, 4(3&4).
[1]Confounding variables are āextraā variable that we didnāt account for; such as the IQ of the students and so on. If not considered, the results could be useless and unacceptable.
