How many studies are in the file drawer? An estimate from the family/marital psychotherapy literature☆
References (28)
Sampling theory
The “file drawer problem” and tolerance for null results
Psychological Bulletin
(1979)Publication bias and meta-analysis
Evaluation in Education
(1980)Directory of clinical members, 1984–1985
(1985)Membership directory
(1986)Directory of the American Psychological Association
(1985)1984 membership directory
(1984)Survey research methods
(1973)- et al.
Systems of family treatment: Substance or semantics?
Annual Review of Psychology
(1988) Statistical power analysis for the behavioral sciences
(1977)
Family dysfunction in bipolar disorder: Description and pilot treatment study
Family therapy pilot study scores on miscellaneous dependent variables
Consequences of prejudice against the null hypothesis
Psychological Bulletin
Research on marital and family therapies
Cited by (39)
An international survey indicated that unpublished systematic reviews exist
2009, Journal of Clinical EpidemiologyCitation Excerpt :The effectiveness of these methods has been verified [17–19]. The survey consisted of 26 questions, which were based on past surveys identified in the literature [4–8], and discussion with experts in survey design, SR methodology, and/or publication bias (Appendix B; available on the journal's website at www.jclinepi.com). To minimize position bias, the order of responses was randomized in the rows of questions for which no order was required, which occurred for eight of the 26 survey questions.
Bias in Meta-Analytic Estimates of the Absolute Efficacy of Psychotherapy
2007, Review of General PsychologyNegative Results: Conceptual and Methodological Dimensions in Single-Case Intervention Research
2018, Remedial and Special EducationA Duty to Describe: Better the Devil You Know Than the Devil You Don’t
2014, Perspectives on Psychological Science
- ☆
This work was supported in part by a grant to the first author from the National Institute of Mental Health (No. 2-R01-MH41097-02). A previous version of this manuscript was presented at the American Evaluation Association, Boston, October 1987.
- 1
The authors would like to thank Jeffrey Berman, Thomas D. Cook, William Follette and Robert Rosenthal for helpful comments on the previous version. We would especially like to thank Martin Frankel for computing confidence intervals for those instances in the present data where the exact binomial probabilities were required.