Four validities

1. statistical

1.1. How well does the data support the claim?

1.2. Is the difference statistically significant?

1.3. How large is the effect?

2. internal

2.1. Was random assignment used?

2.2. Are there any alternative explanations?

2.3. Are there any design confounds?

3. construct

3.1. How well did the researchers measure the dependent variable?

3.2. How well did the researchers operationalize the independent variables?

4. external

4.1. Can the causal relationship generalize to other people, places and times?

4.2. How were the participants recruited?

4.3. Was random sampling involved?