Comm 308 Midterm 2 - Surveys Pt 2
Terms
undefined, object
copy deck
- Interview Mistakes include:
-
- too long: shouldn't be more than 20-30 minutes because of respondent fatigue (giving easy answers at the end)
- too quick to fill in answers: silence makes people think
- taking notes when a recorder would be better
- being judgmental either positively or negatively - Interview Protocols:
-
- funnel: broad to narrow - is best for kids so they get used to questions and help them focus
- reverse funnel: narrow to broad - good when you think you'll run out of time
- hourglass - 3 types of followup questions:
-
- mirror - repeat what they say in question form for more info
- probing question - asking for elaboration - "What do you mean by that?"
- climate question - "you seem uncomfortable, what is wrong?" - Issues in validity and reliability in questionnaires
-
- literacy level (one grade lower for children, no higher than 8th for parents)
- self-administration is good to set pace, but should be read aloud for people who have trouble
- avoid leading questions
- avoid double barreled questions
- avoid questions that elicit socially desirable response - Pretest vs. Pilot
-
- pretesting is initial testing to make sure we are asking what we think we are asking - can be done with colleagues
- piloting is a walk through - should be done with a similar sample - Issues in picking type of questionnaire
-
- cost
- privacy
- sensitivity
- decision tree complexity
- response rate (phone > mail > time use diary)
- speed/efficiency
- difficulty of reaching population - Telephone Questionnaires
-
- cheaper than F2F, no bias with automated voice, can have open-ended questions, can rephrase questions
- only people with landlines, difficult with children, people hang up, DNC list - Computer Driven/Internet Based
-
- can be visually pleasing, people can go at own pace, can take whenever, can have complex decision branches
- need a computer/internet -
- most anonymous, user-friendly, can go at own pace, whenever/wherever
- literacy, low response rate, costs more, selection bias (people who send it back are most motivated either good or bad) - Face to Face
-
- can go in depth, people won't walk out, can restructure questions, can judge affect
- not good for sensitve topics, timely, costs a lot, have to train people - Computer-Assisted
-
- can use media, go at own pace
- have to bring computer - Time Use Diaries
-
- used to capture how people use and experience time
- useful because of level of detail
- can get duration, enjoyment, context
- bad if people fill it out at the end of the day, we don't spread out usage equally (e.g. cram before an exam), low response rate - Time vs. Task Oriented in Time Use Diaries
-
Time: fill it out in 15 minute intervals
Task: fill it out by task - Alternatives to Time-Use Diaries
-
- beeper method
- monitor media usage by login/logout
- observational study