By NinaBergner - 2/2/2021
Hello!
I have a question regarding uploading a study. I have in total 9 different conditions/groups in my experimental study. Every participant should take only once part. When uploading the study to the millisecond website, I should indicate:' choose how participants ids should be generated',' choose how the group id should be determined', and 'enter the total number of groups if any in the experiment'. As I want to have an even number of participants in each condition, I indicated for the first and for the second question: Sequential. For the number of groups, I chose 9.
However, when starting the study a couple of times, I recognized that the study is not sequential, it seems to be random. That's why I am afraid that the selection of participants to conditions will be not even. Should I indicate something else, such as random with replacement? Or does it make no difference?
Best,
Nina
|
By Dave - 2/2/2021
+xHello! I have a question regarding uploading a study. I have in total 9 different conditions/groups in my experimental study. Every participant should take only once part. When uploading the study to the millisecond website, I should indicate:' choose how participants ids should be generated',' choose how the group id should be determined', and 'enter the total number of groups if any in the experiment'. As I want to have an even number of participants in each condition, I indicated for the first and for the second question: Sequential. For the number of groups, I chose 9. However, when starting the study a couple of times, I recognized that the study is not sequential, it seems to be random. That's why I am afraid that the selection of participants to conditions will be not even. Should I indicate something else, such as random with replacement? Or does it make no difference? Best, Nina
Your study is assigning conditions based on subject number per
/ subjects = (1 of 9) to /subjects = (9 of 9)
in your <expt> elements, so the settings for the group ID are irrelevant. Your script does not use group number to assign conditions.
If you wanted to assign based on group number, you would have to specify
/ subjects = (1 of 9)
/ groupassignment = groupnumber
to
/ subjects = (9 of 9)
/ groupassignment = groupnumber
Subject numbers are generated sequentially, so if you visit the study's start page, hit start, finish the study, hit start again, finish the study, hit start again, you'll get sequential numbers and hence sequential conditions unless somebody else also visits the start page during that time and thus consumes one of the intervening IDs. The same is true for the group numbers, although as noted above, they are not actually used in your current setup.
Note that IDs (both subject and group) are consumed whenever somebody visits the study's start page, regardless of whether that person ends up actually launching and participating in the study at that time. This is why you'll see gaps between the IDs in your data. Given a large enough number of participants, those should approximately even out across conditions, but you probably won't get an *exactly* equal amount of participants in each condition.
|
|