This project has moved. For the latest updates, please go here.

Data import, formatting data

Jun 25, 2012 at 1:21 PM
Edited Jun 25, 2012 at 1:22 PM

Hello,

I am trying to import facelab data in OGAMA. However, I am having trouble using the import assistant.

I am thinking of restructuring my data to fit the import assistant. However, I do not understand how to format some of the data. Is there a guide somewhere that will explain the expected structure of the log file?

 

One of the terms is not clear to me. What is the definition of a trial? Is this the showing of a single slide? Hence, if I have 20 slides with gaze points there are 20 trials?

What are:

  • TrialSequence
  • TrialID
  • TrialCategory

 

Furthermore, how do I need to format time and coordinates in step 2?

  • Time (is this in seconds or miliseconds? Is it POSIX time? Or time elapsed since the start of the experiment?
  • GazePosX and GazePosY (Is this expressed in pixels?)

 

I understood from another codeplex discussion that I have to place at the start of every trial (if I understood the concept of trial correctly) the following in the log file: #MSG IMAGE_NAME.JPG

  • Does this line have to contain more than the image file name? Such as time?
  • Where should the images be placed?
  • If I use the above MSG line will the trial start time then automatically calculated (step 4 of the import assistant)?

 

Thanks in advance,

David

Coordinator
Jul 1, 2012 at 8:09 PM

Hi David,

will try to answer your questions ...

> I am thinking of restructuring my data to fit the import assistant.
> However, I do not understand how to format some of the data. Is there a guide somewhere that will explain the expected structure of the log file?
Yes here is my guide :-).
Think of a list with at minimum three columns: time stamp in ms, gazeposx, gazeposy in screen coordinates. In between the lines there may be trigger lines like : "MSG newImage.jpg" or you might have a fourth column containing id (slide number) and fifth column with image file.

> One of the terms is not clear to me. What is the definition of a trial? Is this the showing of a single slide? Hence, if I have 20 slides with gaze points there are 20 trials?
By default you have multiple trials each containing one slide (so no difference between slide and trial at least). In rare cases one would like to have an array of slides aggregated into an analysis unit (e.g. replay doesn't stop after one slide but after all slides in that trial). So in your case 20 slides are 20 trials (each containing on of the slides).

>What are: TrialSequence, TrialID, TrialCategory
TrialSequence gives the order the slide/trial is shown to the subject. TrialID is the unique identifier for a slide. TrialCategory is an optional column to group trials (e.g. angry slides, happy slides). In most cases trialsequence and trialID will be the same, but think of slideshows where some slides reoccur or are randomized for each subject.

> Furthermore, how do I need to format time and coordinates in step 2?
> Time (is this in seconds or miliseconds? Is it POSIX time? Or time elapsed since the start of the experiment?
Time can be seconds, millisecond or mircoseconds, just specify in the import dialog. It can be posix time or time since beginning, Ogama will convert the into time since start during import.

> GazePosX and GazePosY (Is this expressed in pixels?)
Yes in pixels. 

> I understood from another codeplex discussion that I have to place at the start of every trial
> (if I understood the concept of trial correctly) the following in the log file: #MSG IMAGE_NAME.JPG
Yes it can but not necessarily as stated above. it can also just be IMAGENAME.JPG

> Does this line have to contain more than the image file name? Such as time?
No the time is not needed, the time is used from the next valid gaze sample after this msg line.

> Where should the images be placed?
In the SlideResources subfolder of your experiment as stated after succesful import in the assistants dialog.

> If I use the above MSG line will the trial start time then automatically calculated (step 4 of the import assistant)?
Yes, from the next valid gaze sample line.

Hope this helps,
Kind regards, Adrian

Jul 2, 2012 at 9:59 AM

Thank you Adrian. I got a lot further and really a lot further than with the facelab software!

I assume I can create any trial category I like and OGAMA will automatically group participants form the same category.

 

By chance, do you know a software to shift gaze points? Somehow my gaze points are shifted when compared to the image. The lower the participant looked on the slide the more the gaze point is shifted. It almost looks like a logarithmic shift. I know this is not caused by OGAMA.

http://www.korna.it/example1.jpg

Jul 2, 2012 at 2:49 PM

Hi Korna,

 

This can be a measure error due to subject movement after calibration. To prevent this from happening a chin rest is advised.

Within the fixation module you can shift the fixations on the x an y axis (linear though).

 

Best,

Erik

 

Coordinator
Aug 27, 2012 at 5:57 PM

Hi Korna, your image indicates a wrong mapping function, so as Erik says, movement after calibration or just miscalibration. This cannot be corrected with ogamas tools, cause this by default heavily varies between subjects and trials. You have to use better calibration or less head movement. The facelab data may also be not correctly transformed from their internal 3D system to 2D screen coordinates.
Kind regards, Adrian