Publication of nonsignificant results: a survey of psychologists' opinions.

A total of 236 Ph.D. and 1 M.A. faculty completed an e-mail survey regarding their opinions about publication of nonsignificant results. Tenured faculty were significantly less likely than nontenured faculty to write manuscripts for studies with nonsignificant results. Reasons for not writing a manuscript included a perceived inability to publish the manuscript, time concerns, flawed methods or design, inability to interpret results, and the belief that the results are unimportant. The presence of the file drawer problem is supported, and possible solutions are discussed.