CODEWITHSUNDEEP
psychopy
The Wall 79
The Wall 79

Reputation: 31

Pseudo randomization constraints

I am programming a task switching experiment with 3 tasks. The aim of the experiment is to investigate sequential effects: triplets in which the task X repeat after a switch (e.g. ABA or CAC) will be compared with triplets in which the task always switch (e.g. CBA or BAC).

To this aim, is important that the 3 tasks never repeat and that in each block there are (roughly) the same number of repeat and switch sequences.

Each block has 108 trials resulting in 106 triplets (the first two trials cannot be classified as repeat or switch of course).

I have tried to find a solution with several softwares (Psychopy, Conan, Excel), but I haven't found a solution, and I have no clue how to do it.

Any help would be much appreciated

Upvotes: 0

Views: 124

Answers (1)

Michael MacAskill
Michael MacAskill

Reputation: 2421

There are only six possible orders in which the task doesn't repeat:

A   B   C
A   C   B
B   A   C
B   C   A
C   A   B
C   B   A

And six where it does:

A   B   A
A   C   A
B   A   B
B   C   B
C   A   C
C   B   C

So to get to 108 trials, you just need to present each of those orders nine times. But this might conflict with your requirement that the tasks don't repeat (but that phrasing is ambiguous and you should be more specific as to the meaning of that constraint.)

Also, phrases like "there are (roughly) the same number of repeat and switch sequences" aren't great when defining an experimental design. Strive for as much precision as possible.

Having said all that, I'm not sure how this is yet an actual programming-related question? You'll need to say exactly what the issue is with implementing this. The programs you mention have wildly different purposes (PsychoPy is for implementing experiments. Excel is not. I don't know what Conan is).

Upvotes: 1

Related Questions