psychopy.org | Reference | Downloads | Github

Between-Subjects Design and Complex Randomization

OS (e.g. Win10): macOS Big Sur (11.4)
PsychoPy version (e.g. 1.84.x): v2021.1.4
Standard Standalone? (y/n) If not then what?:
What are you trying to achieve?:

I am trying to randomize participants into numerous groups and randomize the stimuli presented within that group. This study is an eyewitness study where a participant views a crime video involving a target individual, and later is presented with a lineup procedure and are asked a few questions about their confidence in their decision. I have 3 target individuals shown in 5 different versions of a video, resulting in a total 15 videos currently being used in the study. A participant is then shown a target-present (person from the video is present) or target-absent lineup (person from the video is absent). While I have one version of the target-absent lineup for each target, there are 5 versions of the target-absent lineup for each target. So, in total there are 18 different lineups. Importantly, 50% of participants need to see a target-absent lineup and 50% need to see a target-present lineup (this 50% need to be subdivided across the numerous target-present lineups corresponding to their target individual).

Is there a way to randomize participants to separate groups and randomize the stimuli presented to them? I usually use Qualtrics and am able to make Embedded Data groups at the start of the study and choose what blocks to present to what participants.

I have had trouble presenting just one video. I have tried creating a two-level loop around the video as described here: Blocks of trials and counterbalancing — PsychoPy v2021.2. However, this method presented all five videos of a target rather than just one. As a result, I felt that I needed to add a third loop where the outer loop contains a list of three Excel sheets-- one for each target which then lists the five different videos as another Excel sheet. The final Excel sheets have the location of one video each. This did not work as I received a “No conditionsFile Found” error.

There is a lot to unpack, so I appreciate help with any one of these issues.
Thank you so much!

If you have spreadsheets or resources defined in code then you need to either add them via Experiment Settings / Online or download resources in code.

Hi, thank you for your response. I suppose one of my struggles is trying to figure out how to structure these spreadsheets. Currently, I am using separate conditions sheets for each routine (i.e. one for the video and one for the lineup task). Is there a more appropriate way to prepare these spreadsheets for a between-subjects design study?

Am I correct in understanding that you have three sets of five videos?

In Qualtrics you could set several embedded variables

target (1, 2 or 3)
video (1, 2, 3, 4 or 5)
lineup (1, 1, 1, 1, 1, 2, 3, 4, 5 or 6)

You should then download and use the appropriate spreadsheets using download resources as per my code snippets

(sorry for the delay – I was on holiday with limited Internet and then my laptop with the partial reply crashed)

Thank you very much!!!