4
$\begingroup$

I'm doing research in Spatial Cognition and am looking for software which will allow me to conduct the following task:

Participants will walk around a space and sit down to a computer task afterwards. There'd be a 2d architectural layout of the space visited by participants presented in the centre of the screen. On the side of the screen, the participants will see randomly scattered image stimuli, representing objects they have seen inside the space. I want them to be able to either drag&drop the images to their assumed locations in the architectural layout, or to click/select an image and click on the place inside the layout where they think they saw a given object. I want to record the final location and the sequence of operations.

I looked at card sorting tools from HCI, but these seem to be too restrictive (won't allow to change background image for example), so I thought someone in this group could hear of psychological software capable of running this sort of task.

|improve this question
$\endgroup$
1
$\begingroup$

This is a Mac related answer - I'm sure there are equivalent applications on other platforms.

When I'm doing usability testing on prototypes we use a combination of OmniGraffle or Keynote and Silverback. Silverback records all the interactions on the screen (will highlight when they click etc.) and will also use the built-in iSight camera to record a video of the user. You can 'mark' interesting points in the recording for easy view later.

On OmniGraffle set the room as a background layer that can't be edited and the objects on another layer.

An alternative is to use something like GoTo meeting and 'broadcast' the session to another room where the team can watch. What you won't get as easily here is highlighting of clicks etc.

Only thing this setup won't do is move on click rather than dragging.

Steve Krug's book "Rocket surgery made easy" is a great getting-started guide to this sort of thing.

|improve this answer
$\endgroup$
  • $\begingroup$ Very useful answer. I should have mention that I'd prefer inexpensive solutions, but your suggestion can be modified by using a full-screen Keynote (or something similar) and a free screencast recorder running in the background. So simple, thanks :) $\endgroup$ – Kuba Krukar Sep 18 '13 at 11:54
  • $\begingroup$ It sounds like this would work well for usability testing. For cognitive psych tasks, I imagine you'd want a means of recording response latencies and user actions in some automated and precise way. $\endgroup$ – Jeromy Anglim♦ Sep 18 '13 at 12:03
  • $\begingroup$ For what I need (only the sequence of moves), this solution would do. Manual coding of the videos will take me less then a day of work and will be 100% reliable (it's pretty obvious from a screencast which image is moved as 1st, 2nd, etc). Writing my own script will take me more than a day (in fact it's been taking me more than a week now:). It's a useful skill for the future, but in general I prefer simpler solutions when measuring latencies to the 3rd decimal place is not justified by the hypotheses. $\endgroup$ – Kuba Krukar Sep 24 '13 at 10:58
  • $\begingroup$ Video editing tools like Final Cut Pro / iMovie have a time stamp and enable you to move one frame at a time to get a very accurate reading of when an event happened. I agree Jeromy's approach is the ideal, if time / budget allow especially with large sets numbers of tests but even in those instances it is often beneficial to pilot it with the 'low tech' approach to make sure the design works before committing to coding. $\endgroup$ – AndrewMinCH Sep 24 '13 at 15:49
2
$\begingroup$

I think a good option would be to use a general purpose language. For example, I once used Visual Basic to program a task that allowed for dragging elements around a display. The IDE for Visual Studio makes it easier to create GUIs. You could readily include a background image and so forth. You could also record location of visual elements and each action that a user takes.

|improve this answer
$\endgroup$
  • $\begingroup$ Thanks, but writing my own program would be the last (i.e. most time-consuming) option. PEBL looks good for this too, I just can't believe there's no ready made solution somewhere out there :) $\endgroup$ – Kuba Krukar Sep 11 '13 at 9:43
2
$\begingroup$

At our institute (at a university), all experiments are custom programmed, mostly in Python, often using PsychoPy, or in MATLAB. We have staff, from post-docs to student assistants, that do nothing but write code (such as packages for R). If you plan to do more than one experiment in your life, I would strongly recommend to learn programming them. You'd be very unhappy with the limitations that any pre-made solutions would impose on your research.

|improve this answer
$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.