Week 10 Flashcards
What are the problems with observation?
Reliability, validity, bias, reactivity
How is reliability a problem with observation?
Are you accurately measuring the behaviour.
How is validity a problem with observation?
E.g. set up camera in subway station for nose picking; did you miss any because back turned? Was that a pick or a scratch?
How is bias a problem with observation and how do you avoid bias with observation?
Threatens construct validity; observer bias - see what they expect to see; observer effects - those being observed change their behavior to match observers expectations.
Blind study, explicit coding, multiple observers.
How is reactivity a problem with observation?
People and animals change their behaviour when being watched; big brother effect.
How does a blind study, explicit coding, and multiple observers help avoid bias with observations?
Blind Study - Observers do not know what condition they are observing; e.g. rats given either placebo or viagra.
Explicit Coding - Very clear operational definitions of behaviours; low change of ‘grey areas’.
Multiple Observers - At least 2 observers work independently assess inter-rater reliability.
How do you avoid reactivity?
Blend in or hide (don’t know they are being observed), habituation (get used to your presence), residual effects (participants no longer there, measure their effects/impacts)
Why are surveys bad?
Everybody lies
Give an example of attitudes not following behaviour.
Chinese couple being served in restaurants during 1930s USA. Survey; 90% no. Travel to each restaurant; 1 said no.
What is important when constructing surveys and polls?
Choosing question formats, writing well-worded questions, encouraging accurate responses.
What are the different types of question formats?
Open-ended questions, likert scale (Easy 1 2 3 4 5 Hard), forced-choice format (do you like class, yes or no?), semantic differential format (review, 5/5 stars).
What are the different types of writing well-worded questions?
Leading questions (do you agree that this is terrible), double-barrelled question (do you enjoy longs walks and taking exams), negatively worded question (double negative), question order
What things can influence a survey encouraging accurate responses?
Giving meaningful responses efficiently, using shortcuts (response sets, acquiescence, fence sitting), trying to look good (social desirable), self-reporting more than they can know, self-reporting memories of events.
What is the classification of a census?
If you sample every member of the population
What is an example of convenience sampling?
Going to the campus library to survey 50 people in 45 minutes.
