Null Effects and Publication Bias in Special Education Research. (August 2017)

• Record Type: Journal Article
• Authors: Cook, Bryan G.; Therrien, William J.
• Is Part Of: Behavioral disorders. Volume 42:Number 4(2017)
• Page Start: 149
• Page End: 158
• Publication Date: 2017-08
• Subjects: experimental design -- policy issues -- interventions; Behavior disorders in children -- Periodicals; Child Behavior Disorders; Comportement, Troubles du, chez l'enfant -- Périodiques; Behavior disorders in children; Trouble du comportement; Trouble affectif; Jeune; Periodicals; Periodicals; 618.9289005
• DOI: 10.1177/0198742917709473
• Languages: English
• ISSNs: 0198-7429

Abstract

Researchers sometimes conduct a study and find that the predicted relation between variables did not exist or that the intervention did not have a positive impact on student outcomes; these are referred to as null findings because they fail to disconfirm the null hypothesis. Rather than consider such studies as failures and disregard the null findings, it is important that these studies be published so that all research is available to the field, which (a) enables valid and complete research syntheses and (b) informs future policy, practice, and research. However, null findings are not always published, leading to the possibility of publication bias, a positively skewed research base, and policy and practices based on incomplete data. This special issue of Behavioral Disorders provides an outlet for methodologically sound studies with null findings. In this introductory article, we provide a context for the special issue by discussing the importance of null findings, the problem of publication bias, and ongoing efforts to publish null findings.