Millisecond Forums

Inquisit IAT data file

https://www.millisecond.com/forums/Topic1431.aspx

By Cli - 3/17/2008

Hi everyone,


I'm an undergradute student and am using the self esteem IAT in my study. I have completed testing and am currently having trouble analysing the data script. I have copied and pasted the results of trials 3,4,6&7 into an Excel file for all my participants as I found it difficult to read the columns in the inquisit file and there were a few pilot tests in amongst the data I wish to use and I don't want that data analysed-however is it possible to use the 'mail merge' procedure with SPSS suggested my on the IAT materials webpage using the excel file as it contains all of the same data minus the practice trials?


Thanks a million&sorry for the long windedness of my comment,


Cli :-)

By seandr - 3/19/2008

Hello,


Which IAT script did you base your own from? If it was based on one of these samples, you can download the spss command file to analyze your IAT data. If all you need is the final D score, Inquisit 3 records the running D score for each trial. You can thus get each participant's final D score from the last row of their data in the column labeled values.d.


Hope this helps,


Sean


By Benja - 9/2/2008

Hi,


I'm sorry but for the life of me I cannot find the column labeled values.d on the attached .dat file.  The script is based on the pictureiat one and I did not muck about with the captured values in any way.  The running D score would certainly be of help!


Much appreciated,


Ben

By seandr - 9/2/2008

Looks like this was generated using an older sample. Try downloading the sample called "IAT template for picture stimuli" from our IAT resources page:


http://www.millisecond.com/download/samples/v3/IAT/default.aspx


You'll have to unzip the files to a folder on your computer, then copy your own picture files to that folder, and change the file names in the script to match the file names of your pictures. This script will record a running D score.


-Sean

By Benja - 9/6/2008

Fantastic, works a treat. 

Is there a way to get Inquisit to write two data files for each experiment, one with ALL the data and one with just the participant's assigned number and their final D Score?  Since I am running a batch file with three different experiments and n=500, combining all the data is likely to become difficult, to say the least.

Thanks,

Ben
By Dave - 9/7/2008

After toying around with some ideas, I've come to the conclusion that currently this is not possible with Inquisit. Which leads me to another feature request:

The ability to define several <data> elements / output files with different specs in a single script. In addition, it would be nice to at least be able to seperately configure data output for <survey> and <expt> elements contained in a single script. I.e.: <data> lets you set the output format for the <expt> part, <surveydata> lets you do the same for the survey part. Currently this doesn't seem to be possible. Any thoughts, Sean? And as we're at it: Any thoughts on this, too?

All the best,
~Dave
By seandr - 9/8/2008

This is definitely on my radar.


Providing a means of customizing the data columns and format for survey data is an obvious one, and I've gotten other requests for this. This isn't trivial to add, but not a huge deal either. I had originally envisioned this as a command on the <survey> element, but putting it on the <surveypage> would allow you the flexibilty to specify the data for each page, which is nice. Likely, I'd put it on both. You would use the <survey> setting as a quick way to set the default for all pages, and if you wanted to override that default, you could do so on a given <surveypage>.


I've also been toying around with ways to easily record a single line of summary data (for example the D scores in the IAT) in addition to the raw reaction time data. My first thought is to create a new element called, for example, <summarydata> that looked similar to <data> but could appear multiple times in the script if multiple summaries were desired.


Regarding the other post - I'll post my reply there. My apologies for not responding sooner, I somehow missed your question on that one.


-Sean



By Dave - 9/8/2008

Great ideas, Sean! Especially the <summarydata> thing. Looking forward to whatever you come up with in the future!

~Dave
By chmo - 7/3/2009

I'm looking to create a summary line at the end of my iat data that has the final D-score. Has this functionality been created? And if so, what do I need to do?

By Dave - 7/10/2009


I'm looking to create a summary line at the end of my iat data that has the final D-score. Has this functionality been created? And if so, what do I need to do?





Afaik, this functionality has not been implemented.


By seandr - 7/10/2009

The summary tag hasn't been created yet. However, there are a couple of options with the existing feature set.


1) If you don't want the raw data at all and just want a summary, you can add the /recorddata=false command to all of the blocks except the block named "summary". This block runs a single trial at the end of the IAT that reports the score to the participant. Remove the /recorddata=false and it will write a single line of data for each subject containing the D scores (and some other stuff).


2) If you want the raw data with a summary line at the end, you can simply remove the /recorddata=false from the summary block. This adds an extra line of data with the final D scores. Note that these scores are already available from the last row of data from the test trials, but using the summary block row might make it easier to extract that row at data analysis time.


-Sean

By Kaarinen - 7/16/2009

Hi,



I'm not sure if I should've opened a new topic to this one, but  post it here for now :). Is the running D score calculated in the same way as in the SPSS script, that is, using the improved scoring algorithm? I'm going to run a ST-IAT, and it seems likely that for my purposes just obtaining the D score would be enough. The ST-IAT is based on the template on this website, unmodified (with the exception of categories and stimulus items, of course :)).



Thanks,


Kaarinen

By seandr - 7/17/2009

Kaarinen,


The ST-IAT sample script records a running, trial by trial calculation of the D score in the data file. To get the final score for a given participant, just check the value of the d score on the last row of data for that participant.


Regards,
Sean

By Kaarinen - 7/22/2009

Ok, so I take it that there's no absolute need for making a SPSS syntax for the ST-IAT, at least for obtaining the D score...

By Admin - 7/23/2009

If all you want is D, then you don't need the SPSS syntax.


-Sean

By dbeaton - 12/7/2009

I'm having problems with the picture IAT as well. I did what you suggested and downloaded the template, unzipped the files to a folder on my desktop, copied my pictures into the folder, and changed the file names in the script to my picture names. For some reason, it still gives me the error message "Unable to load the picture....Verify that the file exists and is of a supported format." The error only occurs for "pic1" and "key," (the first ones on each list)  and when I changed them to different pictures, it gave me the same error for the different first items. Does this mean that it's not the pictures themselves that are problematic, and I did something else wrong? Or are ALL my pictures messed up and the error message just tells me about whichever ones are first on the list? Either way, I can't figure out what I'm doing wrong.


<item targetA>


/1 = "pic1.jpg"


/2 = "pic2.jpg"


/3 = "pic3.jpg"


/4 = "pic4.jpg"


/5 = "pic5.jpg"


/6 = "pic6.jpg"


/7 = "pic7.jpg"


/8 = "pic8.jpg"


</item>



<item targetB>


/1 = "key.jpg"


/2 = "ball.jpg"


/3 = "pencil.jpg"


/4 = "mug.jpg"


/5 = "cloud.jpg"


/6 = "lamp.jpg"


/7 = "table.jpg"


/8 = "tree.jpg"


</item>




By Dave - 12/7/2009

First make sure that all your picture files are indeed located in the same folder as your script. Second, double-check that the JPGs are encoded in RGB colorspace -- JPGs with CMYK color mode will not work properly (http://www.millisecond.com/forums/Topic3586.aspx).


Best wishes from a fellow Inquisit user,


~Dave

By Emily - 12/22/2009

Hi Sean,


We had the same problem as the user above, where the picture IAT (the inquisit 2 download version) doesn't provide any d measures. Is it possible to recover this data from a different script?


Thanks for your help.


Emily

By Dave - 12/28/2009

Hi Emily,


obviously I'm not Sean, but I hope you don't mind me responding to your question anyway. After all it's been hanging out unanswered for a few days now. The first thing to clarify is what exactly you mean by 'recovering data'. You're currently using one of the older IAT templates written for Inquisit 2. Since version 2 didn't have the capability to compute the IAT's D measure on the fly, data generated / collected with this script will indeed not contain the D measure -- Inquisit just can't do that. However, you may compute the D measure from the recorded latency data at any time with a suitable stats package. A SPSS command file for the older V2 IAT templates is available right here: http://www.millisecond.com/download/samples/v2/shortiat/shortiat.sps. Tony Greenwald also provides a SPSS syntax file at this site, which you should also be able to use (a few minor modifications might be necessary, though): http://faculty.washington.edu/agg/IATmaterials/Generic%20IAT.29Dec07.zip. Last not least, SPSS syntax for the newer, V3 IAT templates is available here: http://www.millisecond.com/download/samples/v3/IAT/IAT.sps.


Now, if you want to extend your existing script to automatically compute and record D when run with Inquisit 3, that is technically possible. You may refer to the V3 IAT templates to see how the computations are implemented and then try to transfer these into your existing V2 script. However, I believe the better option is to start with one of the available V3 templates (which has all the necessary computational stuff already in place) and recreate your IAT on that basis, i.e. enter your specifice stimulus items, instruction texts, etc. Hope this helps.


Best wishes from a fellow Inquisit user,


~Dave

By Emily - 1/4/2010

Fabulous. Thanks very much.