Has analytical flexibility increased in imaging studies of bipolar disorder and major depression?

There has been extensive discussion of problems of reproducibility of research. Analytical flexibility may contribute to this, by increasing the likelihood that a reported finding represents a chance result. We explored whether analytical flexibility has increased over time, using human imaging studies of bipolar disorder and major depression. Our results indicate that the number of measures collected per study has increased over time for studies of bipolar disorder, but not for studies of major depression.