Question: Define 'internal validity' based on the provided text. In your answer, explain why experimental designs typically achieve high internal validity while non-experimental correlational designs exhibit low internal validity, referencing the specific examples mentioned in the text.

Sample answer: Internal validity is the degree to which an empirical study's design justifies the conclusion that changes in the independent variable directly caused the observed differences in the dependent variable. Experimental designs achieve high internal validity because they involve deliberately manipulating an independent variable while controlling extraneous variables, thereby eliminating alternative causal explanations. For example, the Darley and Latané experiment manipulated the number of perceived bystanders to isolate it as the cause of helping behavior. In contrast, non-experimental correlational designs exhibit low internal validity because they do not manipulate variables. For instance, a correlational link between exercise and happiness cannot confirm causation because of potential reverse causation or other factors.

Key points:

• Internal validity indicates the degree to which a study's design justifies a causal conclusion between the independent and dependent variables.
• Experimental designs achieve high internal validity by deliberately manipulating the independent variable and controlling extraneous variables.
• The Darley and Latané bystander experiment is an example of high internal validity through rigorous manipulation.
• Non-experimental correlational designs exhibit low internal validity because variables are measured but not manipulated.
• A statistical relationship in a correlational design (e.g., exercise and happiness) cannot confirm causation due to alternative explanations like reverse causation.

Rubric: To earn full points, the student must define internal validity in terms of justifying causal conclusions between independent and dependent variables. They must explain that experimental designs have high internal validity due to manipulation and control (citing the bystander experiment), and that correlational designs have low internal validity due to the lack of manipulation and control (citing the exercise-happiness example).