Implementing Your Experiment Flashcards
1
Q
Implementing Your Experiment
A
Leave nothing to chance- plan everything
Expect the unexpected
If something can go wrong, it will
Make sure mistakes won’t compromise study/data -> safeguards
2
Q
After experiment is designed, but before data collection starts:
A
After each step, re-evaluate design; redesign if necessary
- Have others critique your design
- weed out poor versions before they are done - Figure out total time to run experiment on each subject (and schedule accordingly)
- Prepare and test potential stimuli
- Write up all procedures (from moment sub enters, to leaves)
- Write up subject instructions and debriefing (decrease bad variance)
- Determine total # subjects
- determine # per condition
- plan to test 10% more subjects - Determine if any type of person should be excluded as a potential subject
- Plan for backing up data immediately and often
- Select experimenters
- Determine criteria to exclude data from final analysis BEFORE beginning to give hyp. fair test (usually little to none actually discarded)
- Determine how data is to be analyzed (try out with sample data) and figure out how to interpret any possible outcome (learn something no matter what)
- Calibrate all equipment and schedule regular recalibrations
- Be subject yourself
- Run practice subjects or pilot