Chapter 6

Question 1

Open ended questions may be a legit approach,

but analysis is complicated and time consuming.

Answer 1

but analysis is complicated and time consuming.

Question 2

forced choice format

Answer 2

have to pick a choice, could be only two options or a scale

Question 3

Often, though not always, forced choice takes the form of a

Answer 3

not always, forced choice takes the form of a
Likert scale (Seven options ranging from “Strongly agree” to “Strongly disagree”)

Question 4

Alternatively, one could use semantic differential format , where the

Answer 4

where the options range between two different adjectives.

Question 5

Leading questions

Answer 5

can compromise your construct validity.

Question 6

double barreled questions

Answer 6

which ask you two things at once…

Question 7

Negatively worded questions

Answer 7

are problematic simply because

they can be confusing for the participant.

Question 8

Sometimes the respondents mess up the results by using various

Answer 8

Respondents may use various shortcuts ( response sets) that ultimately could screw up the results and conclusions.

Question 9

One response set is

Answer 9

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.

Question 10

One way to safeguard against acquiescence is to use

Answer 10

some reverse worded questions

Question 11

Fence sitting , whether out of laziness or

commitment issues, is another response set

Answer 11

Solution? Get rid of the neutral option.

Question 12

Other potential problems besides response sets:

Answer 12

•People don’t always know the answers.
•The unreliability of memory
•Faking good (social desirability) or faking bad ,
either intentionally or not.

Question 13

Possible ways to deal with faking good:

Answer 13

•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

Question 14

Some research topics/variables aren’t ideally suited to

self report.

Answer 14

ask people about how much they talk, but an

observational approach would be a better fit here.

Question 15

As we know, observational research involving the measurement of behaviour requires a demonstration of

Answer 15

interrater reliability

Question 16

observer bias

Answer 16

wherein observers’ expectations influence their scoring/assessment.

Question 17

In 1963, Rosenthal and Fode

Answer 17

gave students rats; they told half the students that their rats were really smart and the other half that their rats were dumb. Note: the rats did not actually differ.

Question 18

observer effects

Answer 18

(also known as expectancy effects ): the observers’ biases influence the actual behaviour/performance of the subjects! (Note: it’s not the observers’ assessment that differs!)

Question 19

Similarly, Rosenthal and Jacobsen (1968) falsely told teachers that

Answer 19

some of the students in their classrooms were intellectual “growth spurters ” (lots of potential) according to a made up

Question 20

… what can be done to avoid these pitfalls in observational research

Answer 20

•Provide rigorous training/instructions to coders, e.g., via
codebooks . That is, make things as consistent as
•Assess interrater reliability (not an infallible approach).
•Use masked / blind design that minimizes the potentially
corrupting information given to the scorers.

Question 21

reactivity

Answer 21

, wherein being observed changes participant behaviour.

Question 22

… what can be done to get around reactivity?

Answer 22

1.Be sneaky: Blend in and make
unobtrusive observations.
2.Be clever: Measure the behaviour’s results /
traces, rather than the behaviour itself.
3.Be patient: Wait for reactivity to dissipate.