Explain statistical power - Essay Example

Task: Statistical power Statistical power is the possibility of attaining accuracy while conducting a study. According to Rubin (146), the term relates the probable instances in which chances of error increases with a decrease in power size. For example, a power value that is below 0.05 signifies that the improbable negative rate for that study should be below that difference. Additionally, the correlation should highlight the possibilities of an error for studies of similar nature. According to Rubin (133), statistical power verifies power calculations by eliminating the possibilities of null hypotheses in studies.

According to Ellis (71), statisticians conduct power analysis prior to experiments in order to anticipate the possibility of a study outlaying accurate results. The objective is estimating the effects of the larger sample size in relation to its significance on the study. The less the power, the higher the chances of the occurrence type II error and vice verse. The two possibilities results from the likeness of an error in data collection (Ellis 60). Consequently, statistical power is the retention of null with accuracy, hence the reason, researchers strive to attain more statistical power when conducting studies.

The components that determine statistical powers in a research are the sample size, the effect size, the alpha level and the power. Sample size refers to the units available for the study. Alpha levels are the chances of possibilities. Power is the odd that will result from the treatment of the sample (Rubin 148). Lastly, effect size balances the strength of a research by eliminating errors in power analyses. Indeed, an analysis of the components prior to the collection of data is essential to facilitating accuracy in power analyses.

Works CitedEllis, Paul D. The Essential Guide to Effect Sizes: Statistical Power, Meta-Analysis, and the Interpretation of Research Results. Cambridge: Cambridge University Press, 2010. Print.Rubin, Allen. Statistics for Evidence-Based Practice and Evaluation. Belmont, CA: Brooks/Cole, 2010. Print.