Experiment Planning

Some things to think about when designing your web experiment.

What kind of stimuli will you present? Text? Images? Sounds? Videos?
How will the subject respond? Filling in a form, choosing a category, choosing a numeric value, pressing a key quickly, ...
What are the different experimental conditions? Real-word/Nonsense-word, English/French, Happy/Angry, Male/Female, ...
Do you need a balanced design (in which each subject gets all combinations of conditions)?
How many stimuli will be necessary? How many will each subject get? Subjects get tired after 15min.
Is a random stimulus order OK, or do you need constraints on the ordering of stimuli (to prevent bias caused by order)?
Is this a within-subject design or between-groups design? Within subjects designs are more powerful, but not always possible.
Is this a repeated measures design (in which each subject gets each condition more than once)?
What data will you need to collect for each subject? Individual responses? Means per condition?
What graphs will you plot to show the results? How to show off your data.
How will you test whether the measured differences in responses due to the conditions were unlikely to have arisen by chance? Which statistical tests would be appropriate?

Word count: . Last modified: 14:11 31-Oct-2016.