Finding The Sample Size For Analytical Studies And Experiments

1349 Words Nov 18th, 2016 6 Pages
Hypothesis test

How to find the sample size for analytical studies and experiments?
State the null and alternative hypothesis.
Choose the statistical test based on the type of predictor and outcome variables.
Choose an appropriate effect size.
Set type 1 (alpha) and type 2 (beta) error.
Use the appropriate table to to look for the corresponding sample size.

Basic Concepts
Hypotheses: Null and Alternative hypotheses
People often look into statistical relationship through the test of significance. It is a procedure by which clinicians collect information and see if it agrees with their initial hypothesis. Hypothesis tests are often used for comparison of two samples from the same pool. While the null hypothesis (H0) agrees both could have come from the same population, the alternative hypothesis (H1) can comes in three different forms.

The above has illustrated H1 in forms of mean, but it can also be used in other forms. You may have noticed that the alternative hypothesis of (1) includes the ones of (2) and (3). There are indeed different types of hypothesis tests for different alternative hypothesis.

Two-tailed and One-tailed alternative hypothesis
Two- tailed
Two-tailed test is used when we are interested in the relationship of both sides, i.e. dealing with an open-ended H1 like (1). If we are using a significance level of 0.05, we are splitting our alpha into two sides. This means we are putting 0.025 on each tail. Say if our test statistics fall into that…

Related Documents