CHAPTER 25 Ten Easy Ways to Estimate How Many Participants You Need 361

Chapter 25

Ten Easy Ways to

Estimate How Many

Participants You Need

S

ample-size calculations (also called power calculations) tend to frighten

researchers and send them running to the nearest statistician. But if you you

need a ballpark idea of how many participants are needed for a new research

project, you can use these ten quick and dirty rules of thumb.

Before you begin, take a look at Chapter 3 — especially the sections on hypothesis

testing and the power of a test. That way, you’ll refresh your memory about what

power and sample-size calculations are all about. For your study, you will need to

select the effect size of importance that you want to detect. An effect size could be

the difference of at least 10 mmHg in mean systolic blood-pressure lowering

between groups on two different hypertension drugs, or it could be having the

degree of correlation between two laboratory values of at least 0.7. Once you select

your effect size and compatible statistical test, look in this chapter for the rule for

the statistical test you selected to calculate the sample size.

The first six sections tell you how many participants you need to provide complete

data for you to analyze in order to have an 80 percent chance of getting a p value

IN THIS CHAPTER»

» ^{Quickly estimating sample size for}

several basic statistical tests»

» ^{Adjusting for different levels of}

power and α»

» ^{Adjusting for unequal group sizes and}

for attrition during the study