Experimental Design

Question 1

What are the stages in preparing an experiment?

Answer 1

1. Identify your objective
2. Formulate hypothesis
3. Choose the variables
4. Choose experimental design
5. Choose the task
6. Recruit participants
7. Run the experiment
8. Perform statistical tests on the data
9. Analyse and interpret the results

Question 2

What makes a hypothesis bad?

Answer 2

Vague: Hypothesis does not predict an outcome

Complexity: Hypothesis can explain any result

Question 3

What are the experimental variables?

Answer 3

1. Control variables
   - variables that need to be controlled or kept constant
2. Confounding variable
   - variables that can alter the outcome of the
   experiment
3. Independent variable
   - characteristics changes to produce different conditions
4. Dependent variable
   - Measure used to test the hypothesis

Question 4

What are the 2 main experimental designs? What are their strength and weakness?

Answer 4

Between Subject Design

Strength:
- No learning effect
- Less fatigue
- Multiple variables can be tested simultaneously

Weakness:
- Needs many participants
- Individual variability
- Assignment bias

Within Subject Design

Strength:
- Need fewer participants
- Less chance of variation

Weakness:
- Carryover effects
- Fatigue effects
- Practice effects
- Order effects

Question 5

What are other experimental designs?

Answer 5

Matched design and Ladder of experimental validity.

Question 6

How to run the experiment?

Answer 6

1. Pilot the experiment
   - Test a prototype of the experiment
   - Fix it
   - participants run through should be identical
2. Explain what the experiment will involve
3. Give standardised information
4. Get informed consent
5. Give any training necessary

Question 7

What is null and alternate hypothesis?

Answer 7

Null: any observed changes in behaviour is due to chance

Alternate: hypothesis you are trying to demonstrate

Question 8

What are the types of data?

Answer 8

Nominal - Labels for variables, no inherent order to the categories

Ordinal - Ordered, clear ordering of the categories

Interval - Has ordering and the differences between the values are meaningful

Ratio - All the properties of an interval variable and also has a clear definition of zero

Question 9

When to use parametric or non-parametric tests?

Answer 9

Parametric: data is interval/ratio(frequencies of occurrence), data is normally distributed, data can be characterised by measures of central tendency

If Between subjects, Independent T-test

Repeated Measures, Paired t-test

Non-Parametric:

Data is ordinal/nominal

If Between subjects, Mann-Whitney test

If Within subjects, wilcoxon signed rank