Hi. I have collected AHP pairwise comparison questionnaires on the sustainability of construction problem but due to some changes in my research design, I am going to convert the results from AHP to something else such as Likert scale.
Could you please give me a clue on how to convert or transform the results from AHP to Likert or any other statistical analysis?
Have you used Likert scale in your AHP pairwise comparison? If not, then you might can take the minimum value (weight) from AHP as "STRONGLY DISAGREE" and similarly your highest AHP weight as "STRONGLY AGREE", then linearly assign the values between to min and max to other available Likert metrics. possibly you need to round the values to match Likert. Hope it works.
Mojtaba,
Thank you for your suggestion. Is there any reference you can suggest to support this process?
Cheers,
Have a look at this paper which are dealing with a similar problem but not exactly Saeed asked.
http://citeseerx.ist.psu.edu/viewdoc/download?doi=10.1.1.127.7976&rep=rep1&type=pdf
http://smj.sagepub.com/content/7/1/3.abstract
and also this one:
http://stats.stackexchange.com/questions/86831/convert-likert-scale-to-pairwise
Hi dear Saeed. If you just want to change the scale of your results to Likert, then go with Mojtaba's idea. Otherwise, from what I noticed, I think you have desirable final weights and you want the equivalent pairwise matrix. I've seen some Excel calculations based on Eigenvector method to achieve final weights. Try to find one of them, put the final weights and extract your pairwise matrix values through back calculation automatically. It's kinda tricky, but it'll work.Hope it helps.
Dear Saeed:
It is important to know if you have already your data or not. AHP and Scales have different properties and their conversion it is not automatically. Scales (Likert Type) are ordinal variables but Saaty's AHP is a ratio variable. The way to correlate them need to support in this essential characteristic of the measure procedure.
Once we know if you are in the design phase or in the analysis phase will let us make a better advice. I will look forward in order to help a little more.
Manuel
Dear Manuel,
Thanks for your advice. Actually, I have already collected the data based on AHP pairwise comparison and I am in the analysis stage.
- Badges
- Science topic
- Similar topics
- Medicine
- Physiology
- Biosignals
- Biomedical Signal Processing
- Statistical Analysis