Good evening,

For my master thesis I am conducting research in which I want to make decompositional comparisons for Multidimensional Scaling with the following objects:
o Rotary
o Lions
o Unicef
o Oxfam
o Amnesty International
o Salvation Army
o Rode Kruis
o Food bank
o War Child

This means that I will have 36 comparisons, which I think is to much for one respondent to answer. I want to split these comparisons to make sure that every respondent only has to fill out approximately 18 comparisons.

Is this statistically possible and which is the best way to do this?

I will use Qualtrics for the questionnaire and SPSS to process the data. I searched on the internet, but I could not find any information on this matter.

Some possible solutions that I am thinking about:

- Splitting the comparisons into two groups, where one half of the respondents gets to answer the first group of 18 comparisons and the other half gets to answer the second group. The comparisons will be split randomly, while making sure that all objects are in both groups at least two or three times.

- Splitting the objects into one group of 4 and one group of 5 objects. Within each group, all possible comparisons will be made between the objects in that group. This means that every respondent only gets to make comparisons between 4 or 5 specific objects (and does not get questions about the other objects).

- Putting all 36 comparisons into Qualtrics and making sure that every respondent randomly gets 18 comparisons. (is this possible in Qualtrics?)

Any other idea is also welcome! I am new to MDS (and to Qualtrics), so I would be very grateful if anyone could help me out with this! (of course I asked my supervisor this question first, but he also did not know what is statistically the most correct way to do this).

Thank you very much in advance!

P.S. If I made any mistakes with my spelling or grammar, I am sorry, but this might be because I'm Dutch