Chapter 6 Flashcards
Open ended questions may be a legit approach,
but analysis is complicated and time consuming.
but analysis is complicated and time consuming.
forced choice format
have to pick a choice, could be only two options or a scale
Often, though not always, forced choice takes the form of a
not always, forced choice takes the form of a Likert scale (Seven options ranging from “Strongly agree” to “Strongly disagree”)
Alternatively, one could use semantic differential format , where the
where the options range between two different adjectives.
Leading questions
can compromise your construct validity.
double barreled questions
which ask you two things at once…
Negatively worded questions
are problematic simply because
they can be confusing for the participant.
Sometimes the respondents mess up the results by using various
Respondents may use various shortcuts ( response sets) that ultimately could screw up the results and conclusions.
One response set is
acquiescence or yea saying. If someone answers
“strongly agree” for everything, that could be true… or they could just be really lazy. Could also be no saying, but that’s less common.
One way to safeguard against acquiescence is to use
some reverse worded questions
Fence sitting , whether out of laziness or
commitment issues, is another response set
Solution? Get rid of the neutral option.
Other potential problems besides response sets:
•People don’t always know the answers.
•The unreliability of memory
•Faking good (social desirability) or faking bad ,
either intentionally or not.
Possible ways to deal with faking good:
•Anonymity (but it has drawbacks) •Include items to catch fakers, such as: •Have friends rate them •Use measures like the Implicit Association Test that measure implicit opinions
Some research topics/variables aren’t ideally suited to
self report.
ask people about how much they talk, but an
observational approach would be a better fit here.
As we know, observational research involving the measurement of behaviour requires a demonstration of
interrater reliability