Millisecond Forums

data analysis

https://forums.millisecond.com/Topic12502.aspx

By Gabriel Ponte - 3/3/2014

Hi,

I read, that the IAT effect is the difference between the second block ("compatible condition") and the fifth block ("incompatible condition"). Is it possible to show this effect in the data or do I have to compute each reaction time (for every single reaction) manually. That would be a gigantic load of work. Please clarify this issue, if you can. 

thanks, Gabriel
By Dave - 3/3/2014

> I read, that the IAT effect is the difference between the second block ("compatible condition") and the fifth block ("incompatible condition")

As stated, this is incorrect. You may want to re-read Greenwald, Nosek & Banaji, 2003 for details.

> Is it possible to show this effect in the data or do I have to compute each reaction time (for every single reaction) manually.
> That would be a gigantic load of work

Why would you have to compute something for each reaction time for every single reaction? I have a feeling there may be some profound misconception re. what the IAT's D-score is / represents. In essence, it is a difference score (scaled difference between mean reaction times in two conditions -- "incompatible mean rt" - "compatible mean rt") and is as such very much akin to Cohen's d effect size measure.

Moreover, all the IAT scripts made available via the Task Library at millisecond.com automatically compute D-scores and record them in the data file. See https://www.millisecond.com/forums/Topic3444.aspx.