Millisecond Forums

Programming-wise you can set this up by using any of the Picture IAT scripts as a starting point. All you need to do is change the attribute <text> elements to <picture> elements. From a theoretical perspective, I struggle with the idea -- what exactly is this supposed to measure? To construct an extreme example: Consider the prototypical Flowers vs Insects IAT. Your target categories are represented by pictures of flowers and insects respectively. What kinds of pictures would represent your attribute categories (good vs bad)? Arguably, most people find flowers pretty and hence 'good', while they find insects ugly and hence 'bad'. This end you up in the absurd situation where you test whether pictures of flowers (target) are more easily associated with pictures of flowers (attribute) than with pictures of insects.

Thank you for your answer.

> In fact I would like to measure a packaging perception according unhealthy/healthy on the same base as Bad/good. I pre-test several pictures of food and categorize them in unhealthy or healthy. So my target categories would be my food products unhealthy and healthy and my stimulus a packaging.I fail to see how this fits the IAT-scheme. What are your two *target* categories? The two *attribute* categories are healthy vs unhealthy, I take it?> Do I have to realize my IAT with only the same picture unhealthy or healthy?Just as there are several items (traditionally: words) representing good vs bad attribute categories, you can of course have multiple items represent your healthy vs unhealthy attribute categories.

thanks a lot for your help!

You're welcome. If you hit any roadblock or have additional questions, don't hesitate to let me know!