What Sample Size Do I Need?

Description:

Sample size determination is a critical statistical component at the planning stage of a study and is increasingly requested by regulatory authorities in a number of disciplines.  This course provides a practical introduction to the objective determination of sample size.  The role of power is explained and how different factors affect sample size determination, primarily in the context of hypothesis testing.  A number of common hypothesis testing situations are considered.  R, SAS or Stata may be used for practical work.

Choosing an appropriate sample size is a common problem and should be given due consideration in any research proposal, as an inadequate sample size invariably leads to wasted resources.  This course aims to give a practical introduction to sample size determination in the context of some commonly used significance tests.

Examples from a scientific background will be used to highlight the problems associated with sample size determination and suggest potential solutions. Practical work will be based around the statistical computing packages R, SAS and Stata.  Formulae and algebraic notation will be kept to a minimum.

Who Should Attend?
Scientists and related who need to address the problem of sample size determination in planning a study. Participants are assumed to have a working knowledge of sampling distributions, confidence intervals and significance tests for both means and proportions.

How You Will Benefit
This course will give you a sound introduction to sample size determination. Opportunity will be given to participants to discuss general issues related to sample size.

What Do We Cover?

• Concepts of significance and power in relation to significance tests

• Sample size determination for one sample, two samples and paired samples for both t-tests (means) and z-tests (proportions)
• Effect measures: difference and ratio
• Practical problems associated with sample size determination and possible solutions
• The role of confidence intervals

• Why it can be difficult to determine sample size.

Choice of Software

Practical work may be done in one of the following statistical packages for the duration of the course: R, SAS or Stata.

Cost:

University of Southampton Staff and Students - £256
Others - £320

Website and registration:

https://store.southampton.ac.uk/short-courses/southampton-statistical-sciences-research-s3ri

Region:

South West

Keywords:

R, SAS, Stata

Related publications and presentations:

R
SAS
Stata

Back to archive...