Millisecond Forums

(1) Survey integration

Your options include

- finding another way to match data between SM and Inquisit (e.g. by logging respondents IP and / or MAC ; for Inquisit see http://www.millisecond.com/support/docs/v3/html/language/elements/computer.htm). The unsophisticated solution is to have your respondents enter some kind of unique code into both Inquisit and SM and match data based on that.

- Inquisit has basic survey capabilities, you should start by checking out the 'demographic survey' script (http://www.millisecond.com/download/library/v3/demographics/demographics.exp) available from the Task Library.

(2) Data analysis

- I am fairly sure that the Single-Target IAT templates available at http://www.millisecond.com/download/library/v3/IAT/default.aspx calculate D automatically.

- Obviously D can also be computed post-hoc from the raw latency data using a suitable stats program. Refer to the literature (specifically Greenwald, Nosek & Banaji, 2003) for instructions. SPSS syntax for the *standard* IAT templates is available from the above IAT page, however you'll probably have to modifiy it for use with Single-Target IAT data.

- General resources regarding the preparation / analysis of raw RT data:

• Lacroix & Giguere (2006), which should be available at http://www.tqmp.org/doc/vol2-1/vol2-1.htm (site isn't loading at the time of writing).


• Ron Dotsch's tutorial at http://web.me.com/rdotsch/tutorials/inquisit2spss/index.html.


• Jeromy Anglims tutorial at http://jeromyanglim.blogspot.com/2010/07/how-to-process-inquisit-raw-data-in.html.

Regards,

~Dave

I checked with SM and they do not have the capability of starting in SM, going to another place (IAT), and then back again to SM. In the last hour I've been reading in the forum about programming the Likert-scale surveys so that it is all one instrument (leaving SM) out of the picture. However, I can clearly see that I am over my head programming-wise! I have a basic 27 question likert-scale teacher beliefs survey... do you have a template I could use? I have a computer-programmer friend who is helping me, but I'm not sure how far that help will go! I see in the forum one template called "Scores" which I downloaded and looked at. I see how to insert my questions and responses into the script, so that wouldn't be hard. This would work I think, however, I'll need questions on separate pages so that participants can't go back and change responses; would that be a pretty easy fix?

Also, my current script is not programmed to do any calculations at the end for each participant (like the SingleTargetIAT script). That is okay for my actual data collection -- I don't need the participants to get a score back, but will that affect me when I want to analyze the data? As I see my test run data right now, I have a lot of latency scores for each participant. Can you direct me to some reading on the data analysis end (i.e. what to do with all those numbers!), so that I can be thinking about that step?

Again, sorry for the simplistic questions, but that's where I am! Can't be shy at this point! I appreciate all of your help.

Jamie

As far as I understand, that's the way to go. I.e., have your participants start with the survey app (e.g. display a simple page containing general info about the research they're about to participate in), let the survey app do its magic and generate a participant id it can handle. After the welcome page, have the survey app forward subjects to the Inquisit experiment (IAT) with the generated id appended, have Inquisit use that id as subject number (can be configured via the Web Script Wizard) and then forward subjects back to the survey for the demographics, etc. stuff.

Best wishes from a fellow Inquisit user,

~Dave

Thanks so much; sorry I hadn't looked around more in the helps provided.

One question: it says in the notes "the way to do this is to have participants start in the survey..." Does the process still work if I want to start with the IAT and go from there to the survey and then end? Or should I have an opening page in the survey package, with say, demographics, link to the IAT and when finished, link back to the survey for final questions?

Thanks again,

Jamie

See if any of these help:

http://www.millisecond.com/support/docs/v3/html/howto/interopsurveys.htm

and particularly

http://www.millisecond.com/forums/Topic3369.aspx

~Dave

Hi, I have my IAT set up to send people to a Survey Monkey survey after they complete the IAT. I see in my IAT "test run" data a participant ID number (I told the Wizard to go with Sequential), but I don't see how that number hooks up with anything in Survey Monkey. What do you recommend? I know there is a survey component with IAT; is that required? I need to be able to look at one piece of data from IAT and match it to that participant's SM responses as well. (I will also check with survey monkey people to see what they say).

Thanks,

Jamie