PSY-311: Surveys

Interviews (structured)

Strict script of questions

Interviews (unstructured)

Conversational, follow-up q’s, more open

Aspects of interviews

• Costly (interviewer training, salary, travel)

• Often qualitative which can be subjective

• Concerned with interviewer bias that affects participant responses and interpretation of participant answers

Phone surveys

• Structured or unstructured

• Easy to get random sample

• Cheaper

• Fewer social desirability concerns

Questionnaires

• Paper or electronic

• Cheapest

• Surface level info., breadth v. depth (interviews)

• Extremely common

Advantages of questionnaires

• Easy to administer (quick to administer and score, lots of info., requires fewer resources)

• Can assess non-observable variables (ex. self-esteem, attitudes, and beliefs)

• Can assess variables that you cannot (ethically) manipulate (ex. sexual behaviour, gender, etc.)

Disadvantages of questionnaires

• Accuracy is often low

• People lack insight (ex. their own weaknesses, friends beliefs)

• Forget previous behaviour (ex. frequency of sex)

• Socially desirable responding (ex. lying)

• If assessing the IV, cannot demonstrate causation

Questionnaires often used when assessing…

• Demographics (ex. sex, age)

• Attitudes and beliefs

• Past behavior

• Unobservable current behavior

• Motivation and emotion

• Personality traits

How to clearly define variable (survey construction)

• Should reflect research q.

• May include different operations of same variable (ex. conscientiousness: tidiness, paying bills, following rules)

How to develop valid survey q’s (survey construction)

• Should be easy to understand

• Avoid technical terms (conditioning, schizophrenia)

• Avoid vague or imprecise terms (“health problems”)

• Avoid long or run-on q’s

• Avoid double barrelled q’s (a q that asks 2 things at once)

• Avoid leading q’s (ex. “are you in favour of eliminating the wasteful expenses of the budget”)

• Avoid negative wording (ex. “are you opposed to not allowing gays and lesbians to legally marry?)

How do participant answer your q’s (survey construction)

Open-ended and closed-ended

Open-ended questions

Allow participants to respond however they want
Advantages: Tends to improve much more info. (people might not actually provide it)
Drawbacks: Need time to decide how to score responses, takes time to analyse

Closed-ended questions

Advantages: Easy to score
Drawbacks: Might neglect important info., and gives a limited # of responses (ex. favourite color), won’t obtain tangential info. (ex. “do you think this park should be built?” — does not assess why they think that)

Rating scales

• Respondents pick number on a scale (ex. satisfied/not satisfied, no pain/extreme pain)

• Likert scale

• Need to consider # of response options (5-7 is sufficient)

• How to label the scales

Anchors

How to label a scale; labels are known as anchors (ex. how much TV do you watch study)

How to assemble a survey (survey construction)

• Organised in a way that’s easy for participants to understand (is it viewable on diff. devices)

• Order matters (q’s can bias responses to later q’s)

• Place more uncomfortable/personal q’s near the end

• Demographic items are rarely biased and thus go last