Independent variable, treatment levels and design
- Choice of indepent variables depends on research question
- Treatment levels : different conditions to which participants are assigned
- Within subjects design vs between subjects design
- Example : Research question : To what extent do different ads messages affect willingness to buy coca
Type of experiment
- True experiment
- Field experiment
- Quasi experiment
True Experiment
- Full experimental control
- Experimental group vs control group
- Groups are equal (randomization or matching)
Complete experimental control
Advantage
- Study something which in real life is impossible
Disadvantages
- Evaluation apprehension
- Manipulation makes situation less realistic
Field experiment
- Natural setting
- More heterogeneous participants than in lab
- Less evaluation apprehension
- Manipulate variables & randomly assign participants
- But : less experimental control : other factors can influence results
Quasi experiments
- Natural setting
- More heterogeneous than in lab
- Less evaluation apprehension
- Use existing groups and/or existing differneces
- But : no random assignment and/or no real manipulation : less experimental control
In sum :
- Experiments take you closer to causality than surveys alone
- Good designallow you to show effects and understand underlying processes