Statistical Power in Meta-Analysis

2006 ◽  
Author(s):  
Guy Cafri ◽  
Michael T. Brannick ◽  
Jeffrey Kromrey
Keyword(s):  
Statistical Power ◽  
Meta Analysis

Power-Enhanced Funnel Plots for Meta-Analysis

2020 ◽  
Vol 228 (1) ◽  
pp. 43-49 ◽  
Author(s):  
Michael Kossmeier ◽  
Ulrich S. Tran ◽  
Martin Voracek
Keyword(s):  
Statistical Power ◽  
Meta Analysis ◽  
Relevant Information ◽  
Selective Reporting ◽  
Graphical Display ◽  
Funnel Plot ◽  
Evidential Value ◽  
Graphical Displays ◽  
Funnel Plots ◽  
Meta Analyses

Abstract. Currently, dedicated graphical displays to depict study-level statistical power in the context of meta-analysis are unavailable. Here, we introduce the sunset (power-enhanced) funnel plot to visualize this relevant information for assessing the credibility, or evidential value, of a set of studies. The sunset funnel plot highlights the statistical power of primary studies to detect an underlying true effect of interest in the well-known funnel display with color-coded power regions and a second power axis. This graphical display allows meta-analysts to incorporate power considerations into classic funnel plot assessments of small-study effects. Nominally significant, but low-powered, studies might be seen as less credible and as more likely being affected by selective reporting. We exemplify the application of the sunset funnel plot with two published meta-analyses from medicine and psychology. Software to create this variation of the funnel plot is provided via a tailored R function. In conclusion, the sunset (power-enhanced) funnel plot is a novel and useful graphical display to critically examine and to present study-level power in the context of meta-analysis.

Replication of the Superstition and Performance Study by

2014 ◽  
Vol 45 (3) ◽  
pp. 239-245 ◽  
Author(s):  
Robert J. Calin-Jageman ◽  
Tracy L. Caldwell
Keyword(s):  
Task Difficulty ◽  
Statistical Power ◽  
Meta Analysis ◽  
A Priori ◽  
Significant Heterogeneity ◽  
Performance Study ◽  
Improve Performance ◽  
Research Designs ◽  
Series Of Experiments ◽  
And Performance

A recent series of experiments suggests that fostering superstitions can substantially improve performance on a variety of motor and cognitive tasks ( Damisch, Stoberock, & Mussweiler, 2010 ). We conducted two high-powered and precise replications of one of these experiments, examining if telling participants they had a lucky golf ball could improve their performance on a 10-shot golf task relative to controls. We found that the effect of superstition on performance is elusive: Participants told they had a lucky ball performed almost identically to controls. Our failure to replicate the target study was not due to lack of impact, lack of statistical power, differences in task difficulty, nor differences in participant belief in luck. A meta-analysis indicates significant heterogeneity in the effect of superstition on performance. This could be due to an unknown moderator, but no effect was observed among the studies with the strongest research designs (e.g., high power, a priori sampling plan).

Abstract P311: Effect of Non-Invasive Imaging on Cardiovascular Risk Factors and Downstream Testing: A Meta-Analysis of Randomized Controlled Trials

Circulation ◽  
2012 ◽  
Vol 125 (suppl_10) ◽  
Author(s):  
Seamus P Whelton ◽  
Khurram Nasir ◽  
Michael J Blaha ◽  
Daniel S Berman ◽  
Roger S Blumenthal
Keyword(s):  
Risk Factors ◽  
Risk Factor ◽  
Statistical Power ◽  
Controlled Trial ◽  
Meta Analysis ◽  
Cardiovascular Imaging ◽  
Significant Trend ◽  
Inclusion Criteria ◽  
Non Invasive ◽  
Randomized Controlled

Introduction: Non-invasive cardiovascular imaging has been proposed as a method to improve risk stratification and motivate improved patient and physician risk factor modification. Despite increasing use of these technologies there remains limited evidence documenting its effect on downstream testing and improvement in risk factor control. Hypothesis: Addition of the EISNER study to a prior meta-analysis will improve statistical power to demonstrate the downstream consequences of non-invasive cardiovascular imaging. Methods: A comprehensive literature search of the MEDLINE database (1966 through July 2011) was conducted. Major inclusion criteria required: 1) randomized controlled trial design, 2) participants with no known history of coronary heart disease or stroke, and 3) comparison of a group provided with results of a non-invasive imaging scan versus those without results. A total of eight trials with 4,084 participants met the inclusion criteria for this analysis. We analyzed the data using a random effects model to allow for heterogeneity. Results: Among imaging groups there was a significant increase in prescribing for statins (RR, 1.15; 95% CI, 1.01–1.32) and a non-significant trend for increased prescription of aspirin (RR, 1.15; 95% CI, 0.97–1.35), ACE/ARB (RR, 1.12; 95% CI, 0.96–1.31), and insulin (RR, 1.54; 95% CI, 0.75–3.18). There was a non-significant trend towards increased smoking cessation (RR, 1.35; 95% CI, 0.88–2.08). For downstream outcomes there was a non-significant increase in coronary angiography (RR, 1.20; 95% CI, 0.92–1.57), but not for revascularization (RR, 0.92; 95% CI, 0.55–1.53). There was no significant effect of imaging on the change in traditional risk factors. Limitations: There remains a limited number of trials in this important area. Therefore, trials included in this analysis use a variety of different imaging modalities and we were not able to pool the results based on appropriate clinical action (intensification at high risk and reduction at low risk). Conclusions: Non-invasive cardiovascular imaging leads to increased statin use, but associations with other downstream treatments and change in risk factors are not statistically significant. Our results highlight the limited amount of data for describing the downstream consequences after CAC testing.

scholarly journals Coex-Rank: An approach incorporating co-expression information for combined analysis of microarray data

2012 ◽  
Vol 9 (1) ◽  
pp. 32-43 ◽  
Author(s):  
Jinlu Cai ◽  
Henry L. Keen ◽  
Curt D. Sigmund ◽  
Thomas L. Casavant
Keyword(s):  
Microarray Data ◽  
Statistical Power ◽  
Meta Analysis ◽  
Rank Aggregation ◽  
Microarray Data Analysis ◽  
Combined Approach ◽  
Linear Modeling ◽  
Modeling Process ◽  
Genome Wide ◽  
Gene Level

Summary Microarrays have been widely used to study differential gene expression at the genomic level. They can also provide genome-wide co-expression information. Biologically related datasets from independent studies are publicly available, which requires robust combined approaches for integration and validation. Previously, meta-analysis has been adopted to solve this problem.As an alternative to meta-analysis, for microarray data with high similarity in biological experimental design, a more direct combined approach is possible. Gene-level normalization across datasets is motivated by the different scale and distribution of data due to separate origins. However, there has been limited discussion about this point in the past. Here we describe a combined approach for microarray analysis, including gene-level normalization and Coex-Rank approach. After normalization, a linear modeling process is used to identify lists of differentially expressed genes. The Coex-Rank approach incorporates co-expression information into a rank-aggregation procedure. We applied this computational approach to our data, which illustrated an improvement in statistical power and a complementary advantage of the Coex-Rank approach from a biological perspective.Our combined approach for microarray data analysis (Coex-rank) is based on normalization, which is naturally driven. The Coex-rank process not only takes advantage of merging the power of multiple methods regarding normalization but also assists in the discovery of functional clusters of genes.

scholarly journals The association between rs16917496 T/C polymorphism of SET8 gene and cancer risk in Asian populations: a meta-analysis

2018 ◽  
Vol 38 (6) ◽  
Author(s):  
Hui-Xia Wei ◽  
Guo-Xiang Tian ◽  
Ju-Kun Song ◽  
Lian-Jie Yang ◽  
Yu-Pei Wang
Keyword(s):  
Cancer Risk ◽  
Statistical Power ◽  
Genetic Model ◽  
Meta Analysis ◽  
Critical Role ◽  
Epidemiological Studies ◽  
Asian Populations ◽  
Online Databases ◽  
Cancer Types ◽  
Diversity Control

Epidemiological studies have demonstrated close associations between SET8 rs16917496 T/C polymorphism and cancer risk, but the results of published studies were not consistent. We therefore performed this meta-analysis to explore the associations between rs16917496 T/C polymorphism and cancer risk. Five online databases were searched. Odds ratios (ORs) with a 95% confidence interval (CI) were calculated to assess the association between rs16917496 T/C polymorphism and cancer risk. In addition, heterogeneity, accumulative, sensitivity analysis, and publication bias were conducted to check the statistical power. Overall, 13 publications involving 5878 subjects were identified according to included criteria. No significant cancer risk was observed in genetic model of SET8 rs16917496 T/C polymorphism in Asian populations (C vs. T: OR = 1.04, 95%CI = 0.88–1.23, P = 0.63%; TC vs. TT: OR = 1.17, 95%CI = 0.96–1.24, P = 0.11%; CC vs. TT: OR = 0.90, 95%CI = 0.60–1.37, P = 0.63; TC+CC vs. TT: OR = 1.11, 95%CI = 0.90–1.38, P = 0.33; CC vs. TT+TC: OR = 0.92, 95%CI = 0.65–1.30, P = 0.63). Furthermore, similar associations were found in the subgroup analysis of race diversity, control design, genotyping methods, and different cancer types. In summary, our meta-analysis indicated that the SET8 rs16917496 T/C polymorphism may not play a critical role in cancer development in Asian populations.

scholarly journals Partial derivatives meta-analysis: pooled analyses when individual participant data cannot be shared

2016 ◽  
Author(s):  
Hieab HH Adams ◽  
Hadie Adams ◽  
Lenore J Launer ◽  
Sudha Seshadri ◽  
Reinhold Schmidt ◽  
...  
Keyword(s):  
Gold Standard ◽  
Statistical Power ◽  
Large Scale ◽  
Meta Analysis ◽  
Scientific Evidence ◽  
Joint Analysis ◽  
Individual Participant Data ◽  
Partial Derivatives ◽  
Individual Participant ◽  
Post Hoc

Joint analysis of data from multiple studies in collaborative efforts strengthens scientific evidence, with the gold standard approach being the pooling of individual participant data (IPD). However, sharing IPD often has legal, ethical, and logistic constraints for sensitive or high-dimensional data, such as in clinical trials, observational studies, and large-scale omics studies. Therefore, meta-analysis of study-level effect estimates is routinely done, but this compromises on statistical power, accuracy, and flexibility. Here we propose a novel meta-analytical approach, named partial derivatives meta-analysis, that is mathematically equivalent to using IPD, yet only requires the sharing of aggregate data. It not only yields identical results as pooled IPD analyses, but also allows post-hoc adjustments for covariates and stratification without the need for site-specific re-analysis. Thus, in case that IPD cannot be shared, partial derivatives meta-analysis still produces gold standard results, which can be used to better inform guidelines and policies on clinical practice.

scholarly journals Estimating the statistical performance of different approaches to meta-analysis of data from animal studies in identifying the impact of aspects of study design

2018 ◽  
Author(s):  
Qianying Wang ◽  
Jing Liao ◽  
Kaitlyn Hair ◽  
Alexandra Bannach-Brown ◽  
Zsanett Bahor ◽  
...  
Keyword(s):  
Effect Size ◽  
False Positive ◽  
Latent Variable ◽  
Statistical Power ◽  
Animal Studies ◽  
Meta Analysis ◽  
False Positive Rate ◽  
Mean Difference ◽  
Positive Rate ◽  
Meta Regression

AbstractBackgroundMeta-analysis is increasingly used to summarise the findings identified in systematic reviews of animal studies modelling human disease. Such reviews typically identify a large number of individually small studies, testing efficacy under a variety of conditions. This leads to substantial heterogeneity, and identifying potential sources of this heterogeneity is an important function of such analyses. However, the statistical performance of different approaches (normalised compared with standardised mean difference estimates of effect size; stratified meta-analysis compared with meta-regression) is not known.MethodsUsing data from 3116 experiments in focal cerebral ischaemia to construct a linear model predicting observed improvement in outcome contingent on 25 independent variables. We used stochastic simulation to attribute these variables to simulated studies according to their prevalence. To ascertain the ability to detect an effect of a given variable we introduced in addition this “variable of interest” of given prevalence and effect. To establish any impact of a latent variable on the apparent influence of the variable of interest we also introduced a “latent confounding variable” with given prevalence and effect, and allowed the prevalence of the variable of interest to be different in the presence and absence of the latent variable.ResultsGenerally, the normalised mean difference (NMD) approach had higher statistical power than the standardised mean difference (SMD) approach. Even when the effect size and the number of studies contributing to the meta-analysis was small, there was good statistical power to detect the overall effect, with a low false positive rate. For detecting an effect of the variable of interest, stratified meta-analysis was associated with a substantial false positive rate with NMD estimates of effect size, while using an SMD estimate of effect size had very low statistical power. Univariate and multivariable meta-regression performed substantially better, with low false positive rate for both NMD and SMD approaches; power was higher for NMD than for SMD. The presence or absence of a latent confounding variables only introduced an apparent effect of the variable of interest when there was substantial asymmetry in the prevalence of the variable of interest in the presence or absence of the confounding variable.ConclusionsIn meta-analysis of data from animal studies, NMD estimates of effect size should be used in preference to SMD estimates, and meta-regression should, where possible, be chosen over stratified meta-analysis. The power to detect the influence of the variable of interest depends on the effect of the variable of interest and its prevalence, but unless effects are very large adequate power is only achieved once at least 100 experiments are included in the meta-analysis.

A Meta-Analysis Comparing Relational and Semantic Models

2013 ◽  
pp. 394-409
Author(s):  
Keng Siau ◽  
Fiona F.H. Nah ◽  
Qing Cao
Keyword(s):  
Statistical Power ◽  
Meta Analysis ◽  
Data Modeling ◽  
Systems Development ◽  
Data Models ◽  
The Past ◽  
Semantic Data ◽  
Semantic Models ◽  
Semantic Data Modeling

Data modeling is the sine quo non of systems development and one of the most widely researched topics in the database literature. In the past three decades, semantic data modeling has emerged as an alternative to traditional relational modeling. The majority of the research in data modeling suggests that the use of semantic data models leads to better performance; however, the findings are not conclusive and are sometimes inconsistent. The discrepancies that exist in the data modeling literature and the relatively low statistical power in the studies make meta-analysis a viable choice in analyzing and integrating the findings of these studies.

Sign in / Sign up

Export Citation Format

Share Document