Data retrieval issue

[jarvis24]

Not sure I see the problem. I'm looking through your byQuestionnaire .csv data file and I see that you have data for the RR_E&I_PerfectMale.que file. In your .exp file it indicates that, in Condition 1, the subject should receive that .que file and not any of the other 3 perfect/imperfect que files. Accordingly the data from those three .que files are assigned missing values. And the data are there for the remaining .que files such as gender, main, personality and such. Or was this a working data file? If so, could you also attach a similar folder but one which contains the missing data that ought not be missing? If you can attach that then I'll compare it with the one you posted previously and see if I can tell what's up. Does that make sense (I hope!)

[jarvis24]

Ahh. OK, I just went through the separate email you sent with data from the workstations. I *think* I have an idea re what happened. I'll preface this with saying I'm just hypothesizing here!

It looks like around noon on January 20th, you made some final changes to the structure of your experiment. I reviewed the file "RR_Data_2_5_2011\Station1\byQuestionnaire\ResumeR eview.csv" (see attached). You'll notice the headers obviously do not match the data. *BUT* if you copy the header and data from your recent trial run (attached above), ALL of the data from 11:52am on January 20th and onward, match this newer header and data. So I created a revised version of your data file from Station 1 (see attached). I've highlighted in orange all the sessions preceding the changes to the structure. I've highlighted all the sessions in green all the sessions following the final change. Finally, I've highlighted in yellow the data from your recent test run. Notice that the green and yellow data match up perfectly. I haven't gone through the remaining station data but I suspect they will follow the same pattern. So basically you can use all the data from all your machines collected from that last change onward with the newest header from your recent test run. Make sense?

The reason the .sav file may not look the same is that it needs to be cleared/deleted if variables are added *after* the first session is run. Variables can not be added or dropped from the .sav file after it's been created. You'll need to use the .csv files in this case.

As for the data files labeled "RR_Personality_Questions_que.csv", those would have been created if anyone had instructed MediaLab to run a single questionnaire from the File menu rather than opting to run the whole .exp file.

Hope that helps!

[wbshreves]

Hey Support,

Thank you for your quick reply and showing me that the data is at least there. I think my question was too confusing and did not get to the real problem, but I could not edit it until now. When I try and get the .sav files to load into SPSS only one questionairre's worth of data shows up: RR_Main_Questions.que. For the other three questionairres per a condition nothing is being loaded into SPSS.

Is there a link between what you did with the Excel file, and how to reprogram the experiment so that it won't do this in the future?

Thank you very much for your help, we really appreciate you looking into our problem,

wbshreves