If you want to get rid of that give the IAT on gender a try at https://implicit.harvard.edu/implicit
I'm willing to bet that just the experience of taking it will make you somewhat uncomfortable.
If you want to get rid of that give the IAT on gender a try at https://implicit.harvard.edu/implicit
I'm willing to bet that just the experience of taking it will make you somewhat uncomfortable.
For what it's worth, I'd be interested in seeing something which accurately assesses personal biases, because it'd help one in determining where/if their reasoning might be being influenced by irrational factors.
When you try it out for a topic where you do have a strong association you'll actually feel directly what it's measuring and why it's relevant.
P.S. Did you really try to answer as fast as possible? If you take your time it won't be able to measure much.
In fact, attempting to answer quickly is what lead me to feel it was largely a trick, because it switched up locations of gender after consistent streaks of a given gender (or other metric) being presented solely in one place. Several times I'd be just about to click something and then think "Hey, what the heck? Why did it suddenly move the answer?". This represents a confounding variable in the experiment, because it no longer strictly measures beliefs, but instead measures a mixture of beliefs and motor ability (ability to react quickly to the changing locations of answers), so the test can't determine if an answer was due to belief or motor ability.
And I think you were reading too much into the random placement. The test just picks a number of words for each category at random and then determines whether it's easier for you to categorize certain things together.
You'll be able to respond faster to 'good & saint' and 'bad & killer' than 'good & killer' and 'bad & saint' because your brain categorizes them differently and that difference in performance is what this test measures.