Complex randomization

[Athina B.]

Dear all,

I'm still new in formR and still learning. The design I'm trying to implement is relatively straightforward when I talk about it, yet, it seems that it's pretty difficult to implement it in the programs I've tried.

I don't have a table to show you because I'm pretty sure that the suggested randomization methods will not work. So bear with me. I have read basically all threads with the word 'randomization' in there and I think my question falls outside of the existing ones.

The general idea is this:

- I have a lot of stimuli (90 in total) from 3 different categories (A, B, and C; 30 stimuli per category).

- I want to randomly pick 2 of the 3 categories, and present together (on the same page) 1 random stimuli from each picked category, WITHOUT replacement.

- So per page, I want to have instructions, and my 2 randomly picked stimuli, from the 2 randomly picked categories. Without replacement (per participant). On the same page I need to have the option for the participants to choose which stimuli they prefer.

- I want to repeat this process X times per participant, let's say 9, so that I get all combinations of my categories (AB, BC, AC) 3 times each, with different stimuli from each category each time.

- In essence, no participant will get all stimuli, but of course, across participants, all stimuli will have been (randomly) chosen.

Since the randomization I've read about is probably not going to work for me, I wondered if making 3 different surveys (for my 3 categories) and linking them in my main survey would work? If yes, how could I do that?

Any help will be appreciated.

Thank you in advance.

Athina

[Ruben Arslan]

Hi Athina,

basically, I'd encourage you to program the stimulus drawing procedure in R entirely.

Let formr execute that R code, store the results for this participant.

Then, use items that are basically blanks and fill in the details from your stored results.

Here is an example that does something like this (but the procedure is a bit simpler).

https://docs.google.com/spreadsheets/d/1jkts0xp1JmFX5t1A34kLswH7ZyWWn8xArKeCH2-pzH4/edit#gid=2037039862

Best,

Ruben