Assessing heterogeneity in meta-analysis: Q statistic or I² index?

Huedo‐Medina, Tania B.; Meca, Julio Sánchez; Martínez, Faustino Cavas; Botella, Juan

doi:10.1037/1082-989x.11.2.193

Cited by 3,224 publications

(2,393 citation statements)

References 42 publications

Supporting

Mentioning

2,251

Contrasting

Unclassified

Order By: Relevance

“…For example, socioeconomic status (SES) was not examined as a moderator of weight management interventions in primary care settings due to the lack of consistent reporting practices for information regarding SES. Future studies should use similar measurements of SES, such as the Hollingshead FourFactor Index of Socioeconomic Status (Hollingshead, 1975). Similarly, articles did not uniformly provide information on participant satisfaction, acceptability of the intervention, and attendance; thus, these factors were not examined as potential moderators.…”

Section: Discussionmentioning

confidence: 99%

Pediatric weight management interventions in primary care settings: A meta-analysis.

2016

View full text Add to dashboard Cite

Objective Although the primary care setting has been recommended as an acceptable environment for pediatric overweight/obesity treatment, a quantitative analysis has not been conducted to determine the effectiveness of pediatric weight management interventions delivered in these settings. Therefore, the purpose of the current study was to conduct a meta-analysis of weight management interventions for youth in primary care settings. Method A literature search using PsycINFO and PubMed was conducted to identify articles published through October 2015. Eighteen studies (3,358 participants) met inclusion criteria; studies included a treatment and comparison group and targeted individuals or families for treatment. Study characteristics were coded, and study rigor of articles was assessed. Results The overall effect size for change in BMI in primary care weight management interventions compared to control groups was small but statistically significant (d = 0.26, 95% CI [0.14, 0.38]). The number of treatment contacts, treatment months, and visits with a pediatrician emerged as significant moderators of outcome, such that BMI reduction was positively related to greater contact. Conclusions In comparison to control conditions, weight management programs in primary care settings can be effective for BMI reduction, suggesting that primary care is a suitable setting for treatment of pediatric overweight/obesity. Additionally, treatments that were longer in duration, included more contacts (in person or phone), and included more contacts by a pediatrician had greater impact on BMI reduction. Future studies should continue to examine other aspects of acceptability and accessibility as well as demonstrate the effectiveness of interventions on improving psychosocial outcomes.

show abstract

Section: Discussionmentioning

confidence: 99%

Pediatric weight management interventions in primary care settings: A meta-analysis.

2016

View full text Add to dashboard Cite

show abstract

“…The Q-statistic was then used to calculate the I 2 index. While a significant Q-statistic informs whether moderation may be present, the I 2 index informs the extent of the heterogeneity (Higgins and Thompson, 2002;Huedo-Medina et al, 2006). I 2 values below 25% reflect low heterogeneity, 25%-50% medium heterogeneity, 50%-75% high heterogeneity, and above 75% extreme heterogeneity (Fazel et al, 2009).…”

Section: Meta-analytic Methodsmentioning

confidence: 99%

“…I 2 values below 25% reflect low heterogeneity, 25%-50% medium heterogeneity, 50%-75% high heterogeneity, and above 75% extreme heterogeneity (Fazel et al, 2009). I 2 values greater than or equal to 25% were examined for the presence of moderators, as this suggests between study variability in effect sizes was greater than expected by chance and this threshold has been used in other meta-analyses (Huedo-Medina et al, 2006). Meta-regressions were run to test moderator variables using a macro provided by Wilson (2010), "MetaReg," which runs a weighted generalized least squares regression (Wilson, 2010).…”

Section: Meta-analytic Methodsmentioning

confidence: 99%

Stigma resistance is positively associated with psychiatric and psychosocial outcomes: A meta-analysis

Firmin

Luther

Lysaker

et al. 2016

Schizophrenia Research

View full text Add to dashboard Cite

“…Confidence intervals (CI) were also reported at the 95% confidence level. A Q statistic was used to determine the presence of heterogeneity (Huedo-Medina et al, 2006) and an I 2 statistic estimated the amount of variance between studies. Heterogeneity was considered inconsequential if the variance between studies (I 2 ) was <25% and was considered substantial if >75% (Higgins et al, 2003).…”

Section: Statistical Analysesmentioning

confidence: 99%

A meta‐analysis of anatomy laboratory pedagogies

et al. 2017

View full text Add to dashboard Cite

The debate regarding anatomy laboratory teaching approaches is ongoing and controversial. To date, the literature has yielded only speculative conclusions because of general methodological weaknesses and a lack of summative empirical evidence. Through a meta-analysis, this study compared the effectiveness of instructional laboratory approaches used in anatomy education to objectively and more conclusively synthesize the existing literature. Studies published between January 1965 and December 2015 were searched through five databases. Titles and abstracts of the retrieved records were screened using eligibility criteria to determine their appropriateness for study inclusion. Only numerical data were extracted for analysis. A summary effect size was estimated to determine the effects of laboratory pedagogies on learner performance and perceptions data were compiled to provide additional context. Of the 3,035 records screened, 327 underwent full-text review. Twenty-seven studies, comprising a total of 7,731 participants, were included in the analysis. The meta-analysis detected no effect (standardized mean difference 5 20.03; 95% CI 5 20.16 to 0.10; P 5 0.62) on learner performance. Additionally, a moderator analysis detected no effects (P 0.16) for study design, learner population, intervention length, or specimen type. Across studies, student performance on knowledge examinations was equivalent regardless of being exposed to either dissection or another laboratory instructional strategy. This was true of every comparison investigated (i.e., dissection vs. prosection, dissection vs. digital media, dissection vs. models/modeling, and dissection vs. hybrid). In the context of short-term knowledge gains alone,

show abstract

Assessing heterogeneity in meta-analysis: Q statistic or I² index?

Cited by 3,224 publications

References 42 publications

Pediatric weight management interventions in primary care settings: A meta-analysis.

Pediatric weight management interventions in primary care settings: A meta-analysis.

Stigma resistance is positively associated with psychiatric and psychosocial outcomes: A meta-analysis

A meta‐analysis of anatomy laboratory pedagogies

Contact Info

Product

Resources

About