4. Selecting the Appropriate Hypothesis Test
Once you are ready to conduct a hypothesis test, you need to identify the appropriate test given your hypothesis and data. There are many different types of hypothesis tests, so this can be a challenging and daunting task. The chart below is designed to help you select the appropriate hypothesis test.
How to use the chart:
1) How many variables are there? Choose the column for statistical inference with one, two, or more than two variables.
i) If the number of variables is not clear to you, look at the first row named "What do we want to know...?", and select the appropriate column given the question that resembles what you are seeking to answer.
2) Which tests should we use? Which statistic is relevant to us? Use the "Types of Test" and "Statistic" rows to answer these questions.
i) Use the "Types of Test" row to identify the types of hypothesis tests that can be used to answer the type of question you are interested in. Some of these depend on the type of data you have, so familiarize yourself with each one.
ii) Use the "Statistic that tells us how confident we are" row to see what statistic you should be focusing on to determine the result of your hypothesis test.
3) Click on the tests to learn more about them.
i) Each hypothesis test in the chart below is hyperlinked to an outside source that walks you through the hypothesis test, from what it is, to how it is calculated, to how to interpret the results.
|
|
Statistical Inference with One Variable |
Statistical Inference with Two Variables: Bivariate Hypothesis Testing |
Statistical Inference with More than Two Variables : Multivariate Hypothesis Testing |
|
What do we want to know, given what we see in our sample? |
What is the value of a variable in the population? |
Is there a statistically significant relationship between two variables? |
Is there a statistically significant relationship between two variables, holding other variables constant? |
|
Types of tests |
Confidence interval around a sample mean Links to an external site. Confidence interval around a sample proportion Links to an external site. |
Difference of means Links to an external site. (t-test) Difference of proportions Links to an external site. (z-test) Bivariate regression Correlation coefficient Links to an external site. |
Multivariate regression |
|
Statistic that tells us how confident we are |
Margin of error Links to an external site. |
P - Value Links to an external site. Confidence Interval |
P-Value Links to an external site. Confidence Interval |