Effect size calculation in meta-analyses of psychotherapy outcome research. (4th May 2018)

• Record Type: Journal Article
• Authors: Hoyt, William T.; Del Re, A. C.
• Is Part Of: Psychotherapy research. Volume 28:Number 3(2018)
• Page Start: 379
• Page End: 388
• Publication Date: 2018-05-04
• Subjects: meta-analysis -- effect size computation -- dependent effect sizes; metanálise -- cálculo do tamanho do efeito -- tamanhos de efeito dependentes; Metaanalyse -- Effektstärkenberechnung -- abhängige Effektstärken; 後設分析 -- 效果值計算 -- 相依的效果值; meta-analisi -- calcolo dell'effect size -- effect size dipendente; Psychotherapy -- Periodicals; Psychotherapy -- Research -- Periodicals; Psychotherapy -- Periodicals; Psychothérapie -- Périodiques; Psychothérapie -- Recherche -- Périodiques; 616.891405
• DOI: 10.1080/10503307.2017.1405171
• Languages: English
• ISSNs: 1050-3307

Abstract

Abstract: Meta-analysis of psychotherapy intervention research normally examines differences between treatment groups and some form of comparison group (e.g., wait list control; alternative treatment group). The effect of treatment is normally quantified as a standardized mean difference (SMD). We describe procedures for computing unbiased estimates of the population SMD from sample data (e.g., group M s and SDs), and provide guidance about a number of complications that may arise related to effect size computation. These complications include (a) incomplete data in research reports; (b) use of baseline data in computing SMDs and estimating the population standard deviation ( σ ); (c) combining effect size data from studies using different research designs; and (d) appropriate techniques for analysis of data from studies providing multiple estimates of the effect of interest (i.e., dependent effect sizes). Clinical or Methodological Significance of this article: Meta-analysis is a set of techniques for producing valid summaries of existing research. The initial computational step for meta-analyses of research on intervention outcomes involves computing an effect size quantifying the change attributable to the intervention. We discuss common issues in the computation of effect sizes and provide recommended procedures to address them.