Equalizing stimuli between participants


Equalizing stimuli between participants

Author
Message
eknowles
eknowles
Associate Member
Associate Member (68 reputation)Associate Member (68 reputation)Associate Member (68 reputation)Associate Member (68 reputation)Associate Member (68 reputation)Associate Member (68 reputation)Associate Member (68 reputation)Associate Member (68 reputation)Associate Member (68 reputation)

Group: Forum Members
Posts: 3, Visits: 9
Hi,
I need to get ratings for about 1000 photographs. Ideally, I'd like to have participants visit an Inquisit Web study and be randomly assigned a certain number of photos (e.g., 30) to rate. The goal is to obtain 15 ratings per photo.
Is there any way to have Inquisit keep track of how many times any given photo has been presented, and to prevent photos from being presented more than 15 times? This would involve Inquisit preserving and tracking information between visits to the study ... so my hunch is that it's not possible. But I thought I'd ask!
Thanks,
Eric


Dave
Dave
Supreme Being
Supreme Being (648K reputation)Supreme Being (648K reputation)Supreme Being (648K reputation)Supreme Being (648K reputation)Supreme Being (648K reputation)Supreme Being (648K reputation)Supreme Being (648K reputation)Supreme Being (648K reputation)Supreme Being (648K reputation)

Group: Administrators
Posts: 9.6K, Visits: 44K
eknowles - Monday, January 14, 2019
Hi,
I need to get ratings for about 1000 photographs. Ideally, I'd like to have participants visit an Inquisit Web study and be randomly assigned a certain number of photos (e.g., 30) to rate. The goal is to obtain 15 ratings per photo.
Is there any way to have Inquisit keep track of how many times any given photo has been presented, and to prevent photos from being presented more than 15 times? This would involve Inquisit preserving and tracking information between visits to the study ... so my hunch is that it's not possible. But I thought I'd ask!
Thanks,
Eric


You are correct in that this is not possible -- due to how Inquisit Web works (locally on the pariticipants device, essentially as if the participant were running Inquisit Lab), a given session has no knowledge of and cannot track what happened in any other session. What you could do is assemble a number of finite sets consisting of 30 photos each that satisfy the constraint (i.e. sets are constructed such that each photo occurs equally often across participants), treat those sets as between-subjects condition, and then assign an equal number of participants to each condition / set.

eknowles
eknowles
Associate Member
Associate Member (68 reputation)Associate Member (68 reputation)Associate Member (68 reputation)Associate Member (68 reputation)Associate Member (68 reputation)Associate Member (68 reputation)Associate Member (68 reputation)Associate Member (68 reputation)Associate Member (68 reputation)

Group: Forum Members
Posts: 3, Visits: 9
Dave - Monday, January 14, 2019
eknowles - Monday, January 14, 2019
Hi,
I need to get ratings for about 1000 photographs. Ideally, I'd like to have participants visit an Inquisit Web study and be randomly assigned a certain number of photos (e.g., 30) to rate. The goal is to obtain 15 ratings per photo.
Is there any way to have Inquisit keep track of how many times any given photo has been presented, and to prevent photos from being presented more than 15 times? This would involve Inquisit preserving and tracking information between visits to the study ... so my hunch is that it's not possible. But I thought I'd ask!
Thanks,
Eric


You are correct in that this is not possible -- due to how Inquisit Web works (locally on the pariticipants device, essentially as if the participant were running Inquisit Lab), a given session has no knowledge of and cannot track what happened in any other session. What you could do is assemble a number of finite sets consisting of 30 photos each that satisfy the constraint (i.e. sets are constructed such that each photo occurs equally often across participants), treat those sets as between-subjects condition, and then assign an equal number of participants to each condition / set.

Got it. Thanks! I've gone ahead and divided the photos into 53 batches. For between-subject assignment to these batches, would I make 53 expt elements with subjects attributes (1 of 53), (2 of 53) ... (53 of 53)?
Dave
Dave
Supreme Being
Supreme Being (648K reputation)Supreme Being (648K reputation)Supreme Being (648K reputation)Supreme Being (648K reputation)Supreme Being (648K reputation)Supreme Being (648K reputation)Supreme Being (648K reputation)Supreme Being (648K reputation)Supreme Being (648K reputation)

Group: Administrators
Posts: 9.6K, Visits: 44K
eknowles - Wednesday, January 23, 2019
Dave - Monday, January 14, 2019
eknowles - Monday, January 14, 2019
Hi,
I need to get ratings for about 1000 photographs. Ideally, I'd like to have participants visit an Inquisit Web study and be randomly assigned a certain number of photos (e.g., 30) to rate. The goal is to obtain 15 ratings per photo.
Is there any way to have Inquisit keep track of how many times any given photo has been presented, and to prevent photos from being presented more than 15 times? This would involve Inquisit preserving and tracking information between visits to the study ... so my hunch is that it's not possible. But I thought I'd ask!
Thanks,
Eric


You are correct in that this is not possible -- due to how Inquisit Web works (locally on the pariticipants device, essentially as if the participant were running Inquisit Lab), a given session has no knowledge of and cannot track what happened in any other session. What you could do is assemble a number of finite sets consisting of 30 photos each that satisfy the constraint (i.e. sets are constructed such that each photo occurs equally often across participants), treat those sets as between-subjects condition, and then assign an equal number of participants to each condition / set.

Got it. Thanks! I've gone ahead and divided the photos into 53 batches. For between-subject assignment to these batches, would I make 53 expt elements with subjects attributes (1 of 53), (2 of 53) ... (53 of 53)?

Yes, you can either set up 53 conditions per 53 <expt> elements, or you could use conditional <include> elements instead: https://www.millisecond.com/forums/FindPost19972.aspx

eknowles
eknowles
Associate Member
Associate Member (68 reputation)Associate Member (68 reputation)Associate Member (68 reputation)Associate Member (68 reputation)Associate Member (68 reputation)Associate Member (68 reputation)Associate Member (68 reputation)Associate Member (68 reputation)Associate Member (68 reputation)

Group: Forum Members
Posts: 3, Visits: 9
Dave - Thursday, January 24, 2019
eknowles - Wednesday, January 23, 2019
Dave - Monday, January 14, 2019
eknowles - Monday, January 14, 2019
Hi,
I need to get ratings for about 1000 photographs. Ideally, I'd like to have participants visit an Inquisit Web study and be randomly assigned a certain number of photos (e.g., 30) to rate. The goal is to obtain 15 ratings per photo.
Is there any way to have Inquisit keep track of how many times any given photo has been presented, and to prevent photos from being presented more than 15 times? This would involve Inquisit preserving and tracking information between visits to the study ... so my hunch is that it's not possible. But I thought I'd ask!
Thanks,
Eric


You are correct in that this is not possible -- due to how Inquisit Web works (locally on the pariticipants device, essentially as if the participant were running Inquisit Lab), a given session has no knowledge of and cannot track what happened in any other session. What you could do is assemble a number of finite sets consisting of 30 photos each that satisfy the constraint (i.e. sets are constructed such that each photo occurs equally often across participants), treat those sets as between-subjects condition, and then assign an equal number of participants to each condition / set.

Got it. Thanks! I've gone ahead and divided the photos into 53 batches. For between-subject assignment to these batches, would I make 53 expt elements with subjects attributes (1 of 53), (2 of 53) ... (53 of 53)?

Yes, you can either set up 53 conditions per 53 <expt> elements, or you could use conditional <include> elements instead: https://www.millisecond.com/forums/FindPost19972.aspx

Thanks again! I've set up my study for web deployment (https://research.millisecond.com/eknowles/face_ratings_web.web), and technically it works. However, the load time for participants is high (several minutes). I assume this is because Inquisit is preloading all of the stimuli, regardless of whether the participants is eventually going to see them. Is there any way around this? Much appreciated.
Dave
Dave
Supreme Being
Supreme Being (648K reputation)Supreme Being (648K reputation)Supreme Being (648K reputation)Supreme Being (648K reputation)Supreme Being (648K reputation)Supreme Being (648K reputation)Supreme Being (648K reputation)Supreme Being (648K reputation)Supreme Being (648K reputation)

Group: Administrators
Posts: 9.6K, Visits: 44K
eknowles - Thursday, January 24, 2019
Dave - Thursday, January 24, 2019
eknowles - Wednesday, January 23, 2019
Dave - Monday, January 14, 2019
eknowles - Monday, January 14, 2019
Hi,
I need to get ratings for about 1000 photographs. Ideally, I'd like to have participants visit an Inquisit Web study and be randomly assigned a certain number of photos (e.g., 30) to rate. The goal is to obtain 15 ratings per photo.
Is there any way to have Inquisit keep track of how many times any given photo has been presented, and to prevent photos from being presented more than 15 times? This would involve Inquisit preserving and tracking information between visits to the study ... so my hunch is that it's not possible. But I thought I'd ask!
Thanks,
Eric


You are correct in that this is not possible -- due to how Inquisit Web works (locally on the pariticipants device, essentially as if the participant were running Inquisit Lab), a given session has no knowledge of and cannot track what happened in any other session. What you could do is assemble a number of finite sets consisting of 30 photos each that satisfy the constraint (i.e. sets are constructed such that each photo occurs equally often across participants), treat those sets as between-subjects condition, and then assign an equal number of participants to each condition / set.

Got it. Thanks! I've gone ahead and divided the photos into 53 batches. For between-subject assignment to these batches, would I make 53 expt elements with subjects attributes (1 of 53), (2 of 53) ... (53 of 53)?

Yes, you can either set up 53 conditions per 53 <expt> elements, or you could use conditional <include> elements instead: https://www.millisecond.com/forums/FindPost19972.aspx

Thanks again! I've set up my study for web deployment (https://research.millisecond.com/eknowles/face_ratings_web.web), and technically it works. However, the load time for participants is high (several minutes). I assume this is because Inquisit is preloading all of the stimuli, regardless of whether the participants is eventually going to see them. Is there any way around this? Much appreciated.

If you've gone for the 53 <expt> approach, then yes, Inquisit will download all files. The way around this is the conditional <include> approach I've linked to in my previous reply: That way only the files applicable to the respective condition will be downloaded.

GO


Reading This Topic


Login
Existing Account
Email Address:


Password:


Social Logins

Select a Forum....






Millisecond Forums


Search