Translate

Your Account

When conducting research studies one of the most overlooked yet important aspects of the planning process is the estimation of the sample size required to obtain a proper result with respect to the study and it’s objectives.

Why sample size estimation is important -

  1. Sample size estimation helps plan the time, cost, effort and other logistics relating to the study.

  2. The higher your sample size the more statistical power (which is basically the probability the test will reject the null hypothesis when the alternative hypothesis is true) you deliver to the statistical tests you conduct. Thus the more likely you are to discover new relationships and phenomena that are out of the ordinary and the more likely you are to make a scientific discovery. Because there is a relationship between statistical power and sample size it is possible to estimate the required sample size for a desire high power and so use time and money efficiently.

  3. Standard error (which is a measure of how accurate an estimate or a prediction is) has a direct relationship with sample size. Therefore if you want to make sure that your estimates and your results are accurate it is important to plan out the sample size with regards to the level of accuracy you want.

  4. Finally, it is not uncommon for even professional scientists and researchers to end up with poor quality results due to a lack of planning with regards to sample size estimation. Thus not estimating the required sample size for your study’s objectives in advance can lead to a waste of money, time, effort and harsh criticism and or difficulty in getting a paper accepted when sent to a journal for the peer review process and publishing.

The following table gives the methods that are used for sample size estimation and the types of studies they would assist in (each method links to the page which deals with steps required to carry it out)-

 

 

Method for sample size estimation

Type of study it is used in

1

Species Accumulation Curve

Biodiversity Inventory Studies

2

Sample size for sufficient statistical power (coming soon)

Studies involving hypothesis testing (coming soon)

3

Sample size for desired level of accuracy (reduction of S.E) (coming soon)

Studies which need results having accurate estimates of values

     

 

About the Author of this Article

References -

Handbook of Biological Statistics by John H. McDonald. http://udel.edu/~mcdonald/HandbookBioStat.pdf