The importance of validity and reliability in your thesis

No matter what your thesis is about, arguing the validity and reliability of your research is a key part. You should describe this in your research methodology. But what exactly are validity and reliability? Why are they so important? Validity and reliability are often stumbling blocks for students. In this article, we explain why they are so important and how you should approach them.

Why are validity and reliability so important?

Studies should be representative of the studied target group or population. By describing the reliability and validity, you indicate how well your chosen research methods have measured something that allows you to gain an insight into the problem you are investigating.

In your research methodology you should describe how you guarantee the validity and reliability of your results. In your discussion, you can then indicate to which extent your chosen research method also measured what you wanted to measure (validity) and to which extent the results would be the same if you carried out the research in the same way with a different group or population (reliability).

Validity and reliability are closely linked. They both say something about the quality of your research. For example, a measurement can be reliable, which means that the result would be the same with a repeat measurement. However, that same measurement is not necessarily valid, as your research may not have provided the insight you expected. In principle, you should always strive for a valid and reliable investigation.

What is validity?

Validity says something about the content: did you measure what you intended to measure? Validity describes to which extent the instrument or method you have chosen for your research is suitable for that purpose. You need to substantiate how you are going to answer your main and secondary research questions, and why the chosen method or instrument is best suited for this purpose. There are different types of validity. The types [A1] are:

• Internal validity: this indicates whether your conclusion is correct compared to what you have investigated and whether the links you have identified have been interpreted. Simply put, internal validity shows whether or not the interpretation of the causal relationship is correct.
• External validity: this is linked to the extent to which your research results can be generalised. Can they be applied to the entire population? Are they representative? External validity mainly has to do with your sample, which must be random.

What is reliability?

Reliability has to do with the extent to which a measuring instrument or method provides reliable information. A reliable investigation is free of accidental errors. If your research population is excessively small, for example, then your results are by definition less reliable and less representative of the entire population. That being said, a study is never 100% reliable.

With the confidence interval you indicate to which extent you can confidently make a statement about the collected research results. However, results are never 100% representative and you can never say with certainty that the value of your sample is representative of the entire population. With each sample you redraw, the averages will differ. The confidence interval indicates between which values a research outcome is likely to fall. We often work with a confidence interval of 95%. This percentage should be included in your sampling. A handy tool that you can use to calculate the sample you need [A2] is the sample calculator.

However, you can also calculate the reliability with SPSS through the so-called Cronbach’s Alpha. A high reliability is indicated by a value of .80 or higher.

How do you substantiate the validity and reliability of your research?

In your research methodology, you should discuss at least the following points:

• Why you have opted for your chosen research method, and which tools you have used and why (semi-structured interviews or surveys, for example).
• How many respondents you have involved in this research. With quantitative research the focus is on the sample size. With qualitative research you also have to argue to which extent your research population will produce the right results. Of course, you also have to describe the sample itself. Check out our blog article about qualitative research for more detailed information.
• How you selected your respondents. For example, did you receive the customer data from the organisation? Or maybe you found the respondents via social media? Is that the right population?
• What your response rate was. How many people did you invite, how many responded and how many eventually participated? If you conducted a survey, also indicate how many respondents completed the survey in the end.
• How did you encourage the population to fill in your questionnaire or survey? People like to win prizes and if you raffle a discount voucher, there is a good chance that you will also increase your response rate.
• How did you process your results? Did you use SPSS or did you transcribe and encode?

Need help with your research methodology? Book an appointment for a free consultation.