According to a new study covered by NYT blogger Gretchen Reynolds, sitting in a messy office might make people more creative. Reynolds summarized a study by Kathleen Vohs and her colleagues at the University of Minnesota. In the study:
...college students were placed in a messy or a neat office [for ten minutes] and asked to dream up new uses for Ping-Pong balls. Those in messy spaces generated ideas that were significantly more creative, according to two independent judges, than those plugging away in offices where stacks of papers and other objects were neatly aligned.
a) Assuming that Vohs and colleagues randomly assigned people to sit in a messy or a neat office, can this study support the causal claim that "sitting in a messy office makes people think of more creative uses for Ping Pong balls?" Apply the three causal rules to this situation.
b) A friend argues, "Maybe all the creative people were just sitting in the messy room. That's why they came up with more creative solutions." What can you say in response?
c) Note that in this study, they had two judges evaluate the creativity of the uses for Ping Pong balls that each person came up with. What questions might you ask about the construct validity of this operationalization of creativity?
a) This study was an experiment, so it can potentially support this causal claim. There is covariance, because being in a messy office goes with more creativity. There was temporal precedence, because people sat in the messy (or neat) office before they came up with creative solutions (or not). And there was internal validity--random assignment evens out any individual differences between the two groups. As long as there were no design conf0unds (for example, both groups sat in their respective rooms for the same amount of time, were exposed to the same experimenters, and received the same creativity instructions), there is good internal validity here too.
b) Your friend's suggestion won't apply here; since they used random assignment to conditions, the most creative people in the sample were probably evenly split between the two room conditions.
c) When coders rate some behavior (such as the creativity of an answer), it's important that the coders have inter-rater reliability. In the original article published online last month in Psychological Science, the authors report the IRR as 0.81, which is acceptable. You might also want to know more about which solutions the raters thought were creative. That is, their ratings were reliable, but were they valid judgments of creativity?