Testing Multivariate Effect Sizes in Multiple-Endpoint Studies

Timm, Neil H.

doi:10.1207/s15327906mbr3404_3

Cited by 3 publications

(3 citation statements)

References 23 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…All of the included studies used established outcome measures of depression, either clinical observer ratings, self‐reports, or both. However, if the individual studies under review produce multiple‐effect sizes by using more than one depression outcome (e.g., BDI, HAM‐D), the effects cannot be considered independent of one another (Timm ; Morris and DeShon ). Therefore, a common method was used in which two outcomes of each individual study (e.g., BDI, HAM‐D) were incorporated in one effect size (Dunlap et al.…”

Section: Methodsmentioning

confidence: 99%

The treatment of chronic depression with cognitive behavioral analysis system of psychotherapy: a systematic review and meta‐analysis of randomized‐controlled clinical trials

et al. 2016

View full text Add to dashboard Cite

BackgroundChronic depression is a severe and disabling condition. Compared to an episodic course, chronic depression has been shown to be less responsive to psychopharmacological and psychological treatments. The cognitive behavioral analysis system of psychotherapy (CBASP) has been developed as a specific psychotherapy for chronic depression. However, conflicting results concerning its efficacy have been reported in randomized‐controlled trials (RCT). Therefore, we aimed at examining the efficacy of CBASP using meta‐analytical methods.MethodsRandomized‐controlled trials assessing the efficacy of CBASP in chronic depression were identified by searching electronic databases (PsycINFO, PubMed, Scopus, Cochrane Central Register of Controlled Trials) and by manual searches (citation search, contacting experts). Searching period was restricted from the first available entry to October 2015. Identified studies were systematically reviewed. The standardized mean difference Hedges' g was calculated from posttreatment and mean change scores. The random‐effects model was used to compute combined overall effect sizes. A risk of publication bias was addressed using fail‐safe N calculations and trim‐and‐fill analysis.ResultsSix studies comprising 1.510 patients met our inclusion criteria. The combined overall effect sizes of CBASP versus other treatments or treatment as usual (TAU) pointed to a significant effect of small magnitude (g = 0.34–0.44, P < 0.01). In particular, CBASP revealed moderate‐to‐high effect sizes when compared to TAU and interpersonal psychotherapy (g = 0.64–0.75, P < 0.05), and showed similar effects when compared to antidepressant medication (ADM) (g = −0.29 to 0.02, ns). The combination of CBASP and ADM yielded benefits over antidepressant monotherapy (g = 0.49–0.59, P < 0.05).LimitationsThe small number of included studies, a certain degree of heterogeneity among the study designs and comparison conditions, and insufficient data evaluating long‐term effects of CBASP restrict generalizability yet.ConclusionsWe conclude that there is supporting evidence that CBASP is effective in the treatment of chronic depression.

show abstract

Section: Methodsmentioning

confidence: 99%

The treatment of chronic depression with cognitive behavioral analysis system of psychotherapy: a systematic review and meta‐analysis of randomized‐controlled clinical trials

et al. 2016

View full text Add to dashboard Cite

show abstract

“…Many of these authors focus on multiple-outcome dependence (such as Berkey et al, 1996) or describe approaches that can deal with any kind of multivariate structure (e.g., Nam, Mengerson, & Garthwaite, 2003). For example, Timm (1999) proposed methods for estimating effects under multiple-outcome dependence, and Bijmolt and Pieters (2001) compared several approaches for dealing with multiple-outcome dependence. The dependence among effect sizes from multiple-outcome studies results from the correlation among the multiple outcomes or measures; thus multiple-outcome dependence is different from the dependence in effects from multiple-treatment studies that we illustrate in this article.…”

Section: Dealing With Dependence In Meta-analysismentioning

confidence: 99%

The Degree of Dependence Between Multiple-Treatment Effect Sizes

Kim

Becker

2010

Multivariate Behavioral Research

View full text Add to dashboard Cite

We examined the degree of dependence between standardized-mean-difference effect sizes in multiple-treatment studies in meta-analysis in terms of the correlation formula provided by Gleser and Olkin (1994) . To explore the impact of group size and the values of the true multiple-treatment effect sizes, we simplified the formula for the correlation in terms of the ratio of group sizes and under conditions of equality of sample and effect sizes. The results showed that the group-size ratio affects the correlation between effects much more than do the values of the effect sizes. A relatively smaller control-group size and large effect sizes of the same sign were associated with stronger dependence. We also showed that ignoring the dependence between individual standardized-mean-difference effect sizes always decreases the precision of differences in mean effects across studies in a meta-analysis. The difference in precision was largest when treatment groups were much larger than the control group, regardless of the size of the effects or the number of studies in the meta-analysis.

show abstract

“…Most research on the analysis of multivariate effect sizes has dealt with the problem of testing the significance of the model assuming the true model is within a set of nested models. See for example, Hedges and Olkin (1985), Raudenbush, Becker, and Kalaian (1988), Gleser and Olkin (1994), and Timm (1999aTimm ( , 1999b. In this article, we develop and illustrate a procedure for testing between two nonnested multivariate effect size models using a procedure developed by Timm and Al-Subaihi (2001) for evaluating the specification of two seemingly unrelated regression (SUR) models.…”

mentioning

confidence: 99%