Thread Tools  
eigsti
Junior Member
 
Join Date: Nov 2011
eigsti is on a distinguished road Posts: 16
tagging random stim location, Posted March 21st, 2014 #1
Hi, I have developed a paradigm presenting a pic of type A and a pic of type B; they are shown on R/L sides, with the type being randomized across trials. We are presenting subjects with an audiofile of Type A or B, and their job is to choose Pic A or B to match (e.g, we want to know if they have matched correctly).
My question: How to get the TAG (type A, B) to get marked in the data output? I have a "response" event where subjects choose side LEFT OR RIGHT; but because the type A or B is randomly assigned to Left or Right, the response side info isn't informative. I think I could do this inelegantly, by printing the audiofile info in the output, but I'd really like to be able to do this as part of the response line.
Any suggestions? I uploaded the experiment package (sorry, it's large).
Attached Files
File Type: zip PoolProsodyIMSL5.zip (79.80 MB, 6 views)
Reply With Quote
MonikaT
Senior Member
 
Join Date: Mar 2008
MonikaT is on a distinguished road Posts: 1,279
Posted March 25th, 2014 #2
I recommend changing the events that have "Immediately after the event is presented" selected under the Input tab within the Event Editor, to end on a 1 ms time limit. For example your event "Sound File Lamb A" has "Immediately after the event is presented" checked. Also, I see you do not have "Record and save response" selected under the Input tab for most of your events. If you want the event's data to be stored you will need to check this option.


Regarding your tags, we do suggest when using tags that you attach them to the stimulus list item. If you go to Tag Sets under the Experiment menu, and double click on a tag set you created, you will see it's attached to a trial.
Last edited by MonikaT; April 1st, 2014 at 09:40 AM..
Reply With Quote
eigsti
Junior Member
 
Join Date: Nov 2011
eigsti is on a distinguished road Posts: 16
changes to task, Posted March 25th, 2014 #3
Hi Monika,
thanks for your response. I adopted your suggested changes, but there is still a problem:
The "response" event can't be tagged with the "correct response" because which response is correct on a given trial is dependent on which location was randomly selected in the StimLocation. Thus, either the LEFT or RIGHT response might be correct; which is correct is in turn dependent on whether the audiofile was an "instrument" or a "modifer" type (specified by tag in the "soundfile" events.

I think this may be so complicated to fix -- I really have to start data collection now -- but if you can find an answer I'd still be interested to learn how to do this. I've attached the newer version of the experiment file - it should still map onto the same set of stimulus files.
Attached Files
File Type: sl5 PoolProsodyIMSL5_v2.sl5 (62.4 KB, 3 views)
Reply With Quote
MonikaT
Senior Member
 
Join Date: Mar 2008
MonikaT is on a distinguished road Posts: 1,279
Posted March 25th, 2014 #4
All the information you need is in the datafile, the values are there in the column for "Audiofile" tag. As long as the relevant events have the checkmark for "Record and save response", then the desired data will be in that column.

The reason that you might see double entries like this, "instrument (event)|modifier (trial)" is because individual events such as "PicInstrumentalLamb" have been tagged. For cleaner results, I suggest deleting the "Audiofile" tag set, then recreating it. Then only tag the stimulus lists and not any of the non-list events.

Also, here is a screencast that gives some tips to use on your data: http://www.screencast.com/t/mROe6OU3Vj
Reply With Quote
eigsti
Junior Member
 
Join Date: Nov 2011
eigsti is on a distinguished road Posts: 16
Posted March 25th, 2014 #5
Hi there, thanks for your response. I think I didn't explain clearly enough the problem before. I know I can have the relevant tag information SOMEWHERE in the datafile -- but if it's not in the same line as the response, then I have to create syntax and do post-processing to pull out the "correct" status for each response. What I was trying to do was to have the TAG from an event be "printed" in the subsequent RESPONSE event, so that in the datafile, both the relevant stimulus (which thing was presented) and the subject's response were in the same line; this would make data analysis a real snap. And, of course,if you're not doing this kind of random location, that is easy to do!
Reply With Quote
MonikaT
Senior Member
 
Join Date: Mar 2008
MonikaT is on a distinguished road Posts: 1,279
Posted March 26th, 2014 #6
I apologize if I have misread your questions. You are correct in that post-processing is likely to be the best answer in this particular case. Our screencast shows an example of post-processing with spreadsheet formulas, for anyone who might need it.
Reply With Quote
Thread Tools
Posting Rules
Forum Jump:
Copyright Cedrus Corporation, 1991-2014.
All times are GMT -7. The time now is 07:53 AM.