Hello,
Apologies if this has been asked before, I wasn't sure exactly what to search for to find the answer.
I have a survey design where participants are randomly assigned to one of 3 conditions, which I have set using embedded data. Then, they will view 4 different videos of two people labelling an object. For each video, either Person A speaks first, or Person B speaks first. So, I actually have a total of 8 videos; two for each object, one where Person A speaks first, and one where Person B speaks first. Right now, each of the 8 videos + associated follow up questions is in it's own block (so Block 1 = Object X, Person A speaks first, Block 2 = Object X, Person B speaks first, Block 3 = Object Y, Person A speaks first, Block 4 = Object Y, Person B speaks first etc etc for 4 objects).
I would like to set up the survey flow such that each participant always see two videos of Person A speaking first, and two videos of Person B speaking first. It doesn't matter which of the 4 objects has Person A or Person B speaking first, or what order, as long as at the end they have seen 2 videos of each. However, they should only see each object only once (so they never see Person A label object X first, then also Person B label Object X first).
Currently, I am using branching according to condition, and then using the randomizer to choose between Person A or Person B speaking first for each 'pair' of videos (see pic). However, I am pretty sure this will lead to each pair being independent, so the total could be 3 videos with Person A first, 1 video with Person B first, or all 4 videos with Person B first etc.
Is there a way that I can 'balance' each branch so that they always see 2 videos with Person A first, and two with Person B first?