Sample size and power calculations

Aim

To calculate sample size and/or power when applicable.

 

Requirements

  • Documentation of sample size and/or power and how these are calculated.
  • A reader has to be able to reproduce your calculations

 

Documentation

– The following variables are required for calculating sample size:
  • Significance (α)
  • Power (b)
  • Relevant effect
  • Standard deviation (SD) of individual changes
– Outcome of sample size calculation.

 

Responsibilities

Executing researcher: To make sure sample size and/or power are calculated.
Project leaders: To calculate the sample size and/or power for the study, potentially in consultation with a statistician.
Research assistant: N.a.

 

How To

Most grant applications require a sample size calculation. This is associated with the level of statistical power required for the tests, which will be calculated on future study data.
Most importantly is to contact our biostatisticians for assistance. See ‘Inleiding in the toegepaste biostatistiek’ of Prof. dr. Jos Twisk for more information.
There are numerous formulas for calculating sample size and/or power [Cohen, 1998]. The differences correspond to differences in study design. From a statistical point of view, a lot of power calculations should come with critical notes. This is due to the fact that it is often difficult to provide a true estimate of the statistics required for each formula. On the other hand, a sample size calculation is required for most grant applications to allow people to complete the necessary parameter values one way or the other. Here you can find an overview of calculation methods.

 

Appendices/references/links

  • Cohen J. Statistical power analyses for the behavioral sciences.Hillsdale, New Jersey: Lawrence Erlbaum, 2nd edn., 1988.
  • Twisk J. Inleiding in de toegepaste biostatistiek. Amsterdam: Reed Business, 2nd edn., 2010.

 

Audit questions

  1. Has a power calculation been undertaken for each of your primary outcome measures?
    1. If so, has the measurement level of the outcomes been taken into consideration?
    2. If not, why not?
  2. In the event of equivalence research, did you take the particular requirements imposed on the power calculations by this design into account?
  3. Did the analysis investigate whether the power calculations in the protocol are based on realistic assumptions?
    1. If not, did you record this (in articles and reports)?
  4. If the actual sample size was smaller than that calculated in advance, how was this dealt with in the analysis?

 

V3.0: 13 Oct 2016: Minor revision
V2.0: 12 May 2015: Revision format
V1.1: 1 Jan 2010: Translation into English and updated
V1.0: 1 May 2005: Minor textual amendments