2. Sample Size Calculation – Basic Formula

Introduction to Sample Size Calculation Training session with Dr Helen Brown, Senior Statistician, at The Roslin Institute, January 2016. ************************************************ These training sessions were given to staff and research students at the Roslin Institute. The material is also used for the Animal Biosciences MSc course taught at the Institute. ************************************************ A spreadsheet to carry out the calculations described in the presentation may be copied here: http://datashare.is.ed.ac.uk/handle/1... ********************************************* *Recommended YouTube playback settings for the best viewing experience: 1080p HD ************************************************ Content: Sample size calculation Calculate it such that there is a good chance of achieving study objectives Depends on several factors: ---Size of difference (?) to detect, minimal effect of interest Eg a 50% decrease in gene expression ---Significance level (a) (eg 0.05) ---Power (ß) = chance of detecting a significant result if true difference is ? (typically 80%, 90%) ---Standard deviation (SD or s) of data (how much variability) Basic formula for sample size - Continuous data Where ? = size of difference, minimal effect of interest a = significance level (eg 0.05) ß = power, probability of detecting a significant result (typically 80%, 90%) s = SD of data Zp = points on normal distribution to give required power and significance Most suitable for data assumed to be normally distributed Non-normal continuous data : ---no ideal formula ---above formula often suitable as it tends to be conservative (due to SDs tending to be large in non-normal data) Z-values are from the Normal distribution