Some Practical Guidelines for Effective Sample Size Determination

Lenth, Russell V.

doi:10.1198/000313001317098149

Cited by 861 publications

(573 citation statements)

References 7 publications

Supporting

Mentioning

545

Contrasting

Unclassified

Order By: Relevance

“…For more complex statistical models observed power may no longer be a simple function of observed effect size -it may involve additional parameters (although, as long as these are estimated from the sample alone, even then observed power should be interpreted with considerable caution). 9 will erroneously suggest that the lower a p value (and therefore the larger an observed effect) the stronger the evidence is in favour of the null hypothesis (Frick, 1995;Hoenig and Heisey, 2001;Lenth, 2001). For significant results high observed power will act to (falsely) strengthen the conclusions that the researcher has drawn.…”

Section: Retrospective Powermentioning

confidence: 98%

“…The most widely reported misapplication of statistical power is in retrospective or post hoc power calculations (Hoenig and Heisey, 2001;Lenth, 2001;Zumbo and Hubley, 1998). A 8 retrospective power calculation attempts to determine the power of a study after data has been collected and analyzed.…”

Section: Retrospective Powermentioning

confidence: 99%

“…Hoenig and Heisey (2001) and Lenth (2001) both suggest test of equivalence as alternatives to standard null hypothesis significance testing when researchers wish to show treatments are similar in their effects. A clear introduction to such equivalence tests can be found in Dixon (1998), and an inferential confidence interval approach that integrates equivalence with traditional null hypothesis significance tests has also been described (Tryon, 2001).…”

Section: Retrospective Powermentioning

confidence: 99%

See 2 more Smart Citations

Understanding statistical power in the context of applied research

Baguley

2004

Applied Ergonomics

110

View full text Add to dashboard Cite

Estimates of statistical power are widely used in applied research for purposes such as sample size calculations. This paper reviews the benefits of power and sample size estimation and considers several problems with the use of power calculations in applied research that result from misunderstandings or misapplications of statistical power. These problems include the use of retrospective power calculations and standardized measures of effect size. Methods of increasing the power of proposed research that do not involve merely increasing sample size (such as reduction in measurement error, increasing 'dose' of the independent variable and optimizing the design) are noted. It is concluded that applied researchers should consider a broader range of factors (other than sample size) that influence statistical power, and that the use of standardized measures of effect size should be avoided (except as intermediate stages inprospective power or sample size calculations).

show abstract

Section: Retrospective Powermentioning

confidence: 98%

Section: Retrospective Powermentioning

confidence: 99%

Section: Retrospective Powermentioning

confidence: 99%

See 1 more Smart Citation

Understanding statistical power in the context of applied research

Baguley

2004

Applied Ergonomics

110

View full text Add to dashboard Cite

show abstract

“…5,6 A pilot of 20 samples has been suggested as sufficient to perform robust power analysis. 7,8,9 The protocol used in the pilot and main studies should be identical. If pilot data are not available, useful information might be retrieved from relevant literature, such as epidemiological studies.…”

Section: Introductionmentioning

confidence: 99%

Power Analysis and Sample Size Determination in Metabolic Phenotyping

et al. 2016

View full text Add to dashboard Cite

Estimation of statistical power and sample size is a key aspect of experimental design.However, in metabolic phenotyping, there is currently no accepted approach for these tasks, in large part due to the unknown nature of the expected effect. In such hypothesis free science, neither the number or class of important analytes, nor the effect size are known a priori. We introduce a new approach, based on multivariate simulation, which deals effectively with the highly correlated structure and high-dimensionality of metabolic phenotyping data. First, a large data set is simulated based on the characteristics of a pilot study investigating a given biomedical issue. An effect of a given size, corresponding either to a discrete (classification) or continuous (regression) outcome is then added. Different sample sizes are modeled by randomly selecting data sets of various sizes from the simulated data.We investigate different methods for effect detection, including univariate and multivariate techniques. Our framework allows us to investigate the complex relationship between sample size, power and effect size for real multivariate data sets. elegans.

show abstract

“…It is important to understand that it is only statistically meaningful if the power computations are carried out before the data are analyzed (Lenth, 2001;Verrill and Durst, 2005). The article of Verill and Durst (2005) demonstrates that as the sample size increases, the power rapidly approaches one.…”

Section: Introductionmentioning

confidence: 99%

Power computations in time series analyses for traffic safety interventions

McLeod

Vingilis²

2008

Accident Analysis & Prevention

View full text Add to dashboard Cite

The evaluation of traffic safety interventions or other policies that can affect road safety often requires the collection of administrative time series data, such as monthly motor vehicle collision data that may be difficult and/or expensive to collect. Furthermore, since policy decisions may be based on the results found from the intervention analysis of the policy, it is important to ensure that the statistical tests have enough power, that is, that we have collected enough time series data both before and after the intervention so that a meaningful change in the series will likely be detected. In this short paper we present a simple methodology for doing this. It is expected that the methodology presented will be useful for sample size determination in a wide variety of traffic safety intervention analysis applications. Our method is illustrated with a proposed traffic safety study that was funded by NIH.

show abstract

Some Practical Guidelines for Effective Sample Size Determination

Cited by 861 publications

References 7 publications

Understanding statistical power in the context of applied research

Understanding statistical power in the context of applied research

Power Analysis and Sample Size Determination in Metabolic Phenotyping

Power computations in time series analyses for traffic safety interventions

Contact Info

Product

Resources

About