Post Your Answer

Dear Martin, you can conduct the reliability test first on random samples. For example take two samples and then analyse those two sets of data. After getting the two results, compare them with each other. If the answer comes the same or even similar like 90% to 100% something say 93%, then you can say that the questionnaire is 93% reliable.   I hope this helps you in your study. Best Wishes.

Hi Martin, for open ended questions, you can create a scale where you can put the values in hierarchical order. For example you can give numbers like 0=0%, 1=10%, 2=20%, … , 10=100%. Once this is done and the answers are given in this manner, then you can use the cronbach-alpha test for the open-ended questions. However, for the mcqs, the better approach is to use the Kuder-Richardson KR-20 test.   Best Regards, Brett

This is really hard to tell just by the given information as it depends on the research objectives as well. Moreover, different techniques and approaches are used for different reliability tests. As for the MCQs, I know that the face validity analysis or content validity analysis is more appropriate but I am not too sure about the others.   Though I can suggest you the experts who can help you better. You can contact the team of Chanakya research or visit their website. I am attaching the link below. I hope you get the right guidance and do your study well. https://www.chanakya-research.com/

Hello Martin, you are correct with one thing that the cronbach alpha test is good for likert questions and not for the nominal and textual answers received  from the questionnaire. Therefore, I'd suggest you to use the principal component analysis before applying the cronbach approach.   Also, please make sure whether your data has a single group or multiple groups. It can also help in choosing the suitable approach for the reliability test. Good luck.

First you need to standardize the data and then, try the method of alpha-cronbach coefficient. Only after that you can test your questionnaire for correlation between questions and other items as well as check the answers and questions for the reliability. You can even use SPSS for the same.

Hello, thank you for your response. The thing that I am confused about is whether Cronbach Alpha can be used for mcqs and open ended or not because I am only sure about it being a reliable test measure for the likert questions. So, please correct me if I misunderstood the method. Also, I’ll look into the SPSS method once I have the answers. Thank you for the suggestion once again.

This method cannot be used for open ended questions. You can try standardizing our open ended questions. Also, you can get in touch with the PhD Guidance website. They may guide you through the questionnaire development in a better manner. https://www.phdguidance.org/

 Hi dear, I agree with Oliver. It is important to standardize the data first, especially the open-ended one. You can do an exploratory answers analysis with the open-ended questions first by choosing a sample data to reduce the original size and then do a closed analysis to be more definitive of the answer.   I hope I have made sense and that it proves to be helpful.  

Thank you so much everyone for your active participation. I am going to use the measures and see which one provides the most suitable result or which one does not. In any case, I will drop by again to tell you the results. Thank you so much.  

It was a pleasure to be a part of this discussion.