# Choosing a Statistical Test

Teachers or students who want the original PPT please tweet @psychopepper

#### Learning Objectives

• To KNOW and UNDERSTAND how to choose the correct inferential statistical test
• To ANALYSE novel studies and APPLY the decision criteria to determine the correct statistical test.
• To CREATE a mnemonic to revise the decision criteria

### Lesson Outline:

1. Can you remember what P.<0.05 are, and what a type one error is?
2. Attempt an exam question of the likelihood of making a type 1 error
3. Revise and if needed take notes on the three levels of data
4. Revise and if needed take notes of experimental designs
5. Revise and if needed take notes on difference vs relationship
6. Take note of the decision diagram to learn off by heart.
7. Use decision criteria in order to decide which test should be used for 6 research studies and justify the answer.
8. Redesign the study in some way that would change which statistical test would be needed,.

#### Content Recap:

Difference: Is the study testing a difference between two groups or a relationship between two things?

Design: If it is a test of difference, which experimental design did it use?

• Unrelated Data: Data that has been taken from studying two groups which are not related to each other (independent groups).
• Related Data: Data that has been taken from studying two groups which are related to each other (repeated measure or match pairs).

Data: What type of measurements did they use?

• Nominal Data: Data that can be put in categories. For example Pass/Fail or Improve/Decline/Stay Same.
• Ordinal Data: Data which shows us who was better than who (the order) but without any specific measurements. For example GCSE Grades, Scale 1-10.
• Interval Data: Data which have standardised and specific measurements. For example GCSE Marks, Time.