chapter 8

Question 1

likert scale

Answer 1

• respondents can indicate to what extent they agree with or endorse a statement
• likert items should be monotone meaning respondents are consistently more likely to agree with the item if they possess the property to a greater degree

Question 2

likert scale construction

Answer 2

• begins with a clear description of the property to be measured
• distinguish dimensions that will be measured with a subset of items that form a subscale
• items are generated for each unique combination of dimensions with each including a specific situation and expression of the property
• item formulation should be short and simple to avoid misinterpretation
  Items should be neutral not suggestive
• response options should be unambiguous, consistent, exhaustive and mutually exclusive

Question 3

cumulative scale

Answer 3

• include items themselves that show consistent ordering, with each item expressing the property more strongly then the previous one
• not used often
• e.g. “i like cats” → “i love cats” → “i worship cats”

Question 4

surveys

Answer 4

• self-report
• sampling is central consideration because who you talk to is going to have fundamental impact on the kind of data you gather
• open ended questions don’t limit the information that respondent choose to give you but are harder to score
• close ended questions are more straightforward and easily scored

Question 5

responding to surveys involves

Answer 5

• question interpretation: vary from respondent to respondent
• information retrieval
• judgment formulation: think and judge the information in the response we are going to make
• response formatting: process of sharing our information
• response editing: decide to change wording of our response till we are happy with it

Question 6

creating surveys

Answer 6

• brief: simple, straightforward wording for questions
• relevant: relates to ethical considerations; don’t ask information that’s not relevant to your research question
• unambiguous: questions that can only be interpreted one way
• specific: each question should ask one thing only; no double-barrelled questions
• objective: avoid questioning that is leading or suggestive toward a specific response; avoid loaded questions

Question 6

ways to administer questionnaires

Answer 6

• in person to groups or individuals
• mail/email surveys
• online/internet surveys
• other technologies

Question 7

sampling bias (error)

Answer 7

• sample is not representative
• you want your research to be generalizable, if it’s not generalizable you have bias in your sample

Question 7

2 concerns of sampling

Answer 7

1. sampling bias
2. response bias

Question 7

non-response bias

Answer 7

• sampling bias due to systematic non-responding
• consistent characteristic among those who don’t respond will introduce error

Question 8

the silent probe

Answer 8

interviewer looks expectant to silently probe answers without implying what to say

Question 9

parts of an interview

Answer 9

• the silent probe
• overt encouragement
• elaboration: “tell me more about that”
• ask for clarification
• repetition: repeat words they give you in a response to gain clarification
• problems: interview bias