Visual Processing Test - Results

Individual Results

Aggregate Results

Thoughts

At the time of writing (3 Jan 2016), the average discrimination sensitivity (ability to differentiate between control images with nothing in them, and signal images with people/animals in them) increased from 0.62 to 0.66 between the before and after blocks, an increase of 0.04. This is a much smaller increase than the ~0.3 increase (~0.55 to ~0.85) in the original study.

This may be due to the nature of this replication, a "throwaway" online test rather than a "serious" sit-down test with a supervisor as in the original study. It may also be due to the number of sessions, 4 rather than the original 16, 4 may be too few to really get confortable with the test. Another cause may be the test images themselves. The test images I chose may not be as easy to recognise after priming as the ones used in the original study.

The criterion shift (a lower criterion indicates a response bias to report the presence of a person) went from -0.26 to -0.32 in the before and after blocks, a decrease of 0.06. The original study had a shift from ~0.05 to ~-0.1, a decrease of ~0.15. This indicates that my test subjects' confidence after priming increased only about a third as much as in the original study. This also agrees with the discrimination sensitivity results.

Overall, as a side project, this was fun to build and relatively successful. The results above tell me that if I or someone were to spend a little more time getting good test images (see guidelines in the materials and methods PDF), this little tool could be an accurate assesment of a person's top-down visual processing ability, and all the implications that go along with it.

Thanks

Thanks for being a test subject. Return to home.