The art of Sample Size Determination: are 5 participants enough for my study?

This one day course on sample size estimation and power calculation is aimed at academics and researchers mostly working in the fields of medical, health and social sciences. This course provides the basics of sample size estimation. It will be of use to those planning to submit a research project, or committed to generating research results that will be taken seriously for publication or by the scientific community. Statistical software STATA will be used in doing the practical examples.

Course Code aossd12012018
Course Date 12th January 2018
Places Available 22
Course Leader Faiza Tabassum

Course Description

It is practically impossible to collect data on an entire population of interest. Instead we examine data from a random sample to provide support for or against our hypothesis. Now the question is: how many samples/participants/data points should we collect? Power analysis allows us to determine the sample sizes needed to detect statistical effects with high probability. Power and sample size estimations are used by researchers to determine how many subjects are needed to answer the research question (or null hypothesis). This one-day course will offer an introduction to sample size and power analysis and will show how to use it simply and effectively to plan the appropriate sample size for a study.

The target audience

This course is for those who wish to gain an understanding of sample size calculation. Estimation of sample size is required at the time of planning a study and usually it is done when a grant proposal is underway. This course briefly highlights the processes involved in calculating the sample size. This course also addresses the issue of the power of the sample. This course is divided into theory and practical examples using STATA.

Course outline

Sample, population, types I & II errors, hypothesis testing, significance, effect size, power analysis, the essential steps to start calculating sample size correctly, worked examples for many study scenarios and a practical session using STATA.

Participants are expected to have a good working knowledge of simple statistical methods, including a good understanding of estimation of parameters including confidence intervals, hypothesis testing and p-values. No familiarity with the software STATA is required.