Lecture 7 - Surveys Flashcards
Common Survey Topics
- Victimization - Self-Report
- Offending
- Attitudes & Beliefs
- How do people perceive the CJS?
- Good for evaluation
General Layout & Format of Questionnaire
- Easy to read
- One question per line
- Keep data entry in mind
- No abbreviation in question
- Consider overall length
- Clear instructions
Question Wording
- Avoid ambiguity - Clear, direct
- Avoid negative wording
Ex. Do you disagree there should be no mandatory minimum sentencing? - Avoid double-barrelled questios
Ex. Do you support hiring more officers and longer sentences for offenders?
Question Wording: Biased Questions
Avoid Leading Questions
- Encourages respondents to respond in certain way
Ex. Don’t you agree that using illegal substances is bad?
Avoid Loading Questions
- Questions assuming the respondent has done or experienced something, or makes a judgement about the respondent
Ex. Have you quit using drugs?
Social Desireability Bias
Tendency to respond in a way that is socially acceptable
Avoid by:
* Frame as “common behaviour”
* State privacy provisions
* Limit value-based language
Response Formats: Open-Ended
Respondents provide answers in their own words
Advantages:
* Detailed, no limits
* Good for complex answers
Disadvantages
* Non-response
* Can be misinterpreted
Response Formats: Close-Ended
Pre-defined list of response options
* Exhaustive & mutually exclusive
Advantages:
* Less non-response
* Quick, easy to answer
* Easy analysis
Disadvantages:
* Limited response options
* Miss nuance in data
(failing to notice or account for the subtle, often crucial details in the data.)
Question Types: Dichotomous
Offers only two possible response options
Ex.
Are you a CRIM major?
* Yes
* No
What is your student status?
* Domestic
* International
Question Types: Multi-Option Checklist
Questions with multiple options:
What is race:
* Asian
* White
* Middle Eastern
* European
* Black/African
* Indigenous
* Other
Levels of Measurement
Nominal:
- Categories with no order or rank
What is your current living situation?
* Family home
* SFU residence
* Roomates
* Other
Ordinal:
- Categories with order and rank
How often do you remember seeing math posters at SFU?
* Never
* Once per week
* 2-3 times per week
* More than 3 times per week
Levels of Measurement
Interval Questions
* 1-5 - Strongly disagree to Strongly agree scale
Ratio Questions
* How many times per week do you drink alcohol?
Complex Question Formats
Contingency questions
* Skip patterns to ensure the right respondents are answering the appropriate questions
Matrix Questions
* Series of questions with the same responses
Assessing question content
- Is question necessary/useful?
- Can it be misunderstood?
- Multiple questions needed or only one?
- Is the question specific enough?
Solutions
* Use existing measures, do not start from scratch
* Pilot test (survey trial run)
Potential Issues
- Floaters
- Fence-sitters
- Telescoping
- Response rates
Consider Question Order
Potential order effects:
* Boredom
* Recall
Self-Report Surveys:
* Interesting hook question first
* Provocative questions second
* Boringq uestions last
* Never start with open-ended questions
Face-to-face interviews
* Boring questions first
* Provocative questions later
Self-administered surveys: Group settings
Advantages:
* Cheaper, quicker than mail
* High response rates
* Privacy, but can receive help
Concerns
* Anonymity
* Voluntary participation
Self-administered Surveys: Mail
Advantages:
* Privacy
* Cost-effective
* Reaches large sample
* No researcher bias/reactivity
Concerns
* low response rates
* No clarification when asked
* Literacy barriers
* Junk mail
* Slow
Self-administered Surveys: Online
Advantages:
* Cost effective
* Reaches large sample
* Efficient data management
* No researcher bias/reactivity
Concerns
* Generalizability concerns
* Challenges for incentives
* No clarification when asked
* Lieracy barriers
* Response rate
Researcher adminstered surveys: Phone
Advantages:
* Low cost
* Multiple attempts possible
* Flexible
* Wide-reaching
Concerns
* No control over environment
* Phone number access is limited
* Low response/completion rates
Researcher Administered Surveys: Face-to-face
Advantages:
* Establishes rapport with participant -> increased data quality
* High response rate
* Can clarify questions
* Researcher is responsive - can probe encouragement, repetition, elaboration, etc
Concerns:
* Interviewer bias/reactivity
* Social desireability bias
* Expensive, time-consuming
* Difficult to access respondents (travel, safety concerns)
How to increase response rates
- Make simple to submit
- Keep survey reasonable length
- Reminder mailings/emails
- Incentives
