Advanced questionnaire design

Question 1

what is Cronbach’s alpha?

Answer 1

a measure of internal consistency

Question 2

What value is considered reliable for Cronbach’s alpha?

Answer 2

Moderate - 0.6-0.7/8
>0.8 is considered very good
>0.95 might indicate issues with experimental design with too much redundancy in it

Question 3

what are the big 5 personality traits?

Answer 3

openness
conscientiousness
extroversion
agreeableness
neuroticism

Question 4

What are you looking for in a questionnaire pilot?

Answer 4

Wording issues
Clarity of questions
distribution of responses
leading questions

Question 5

what are the causes of all mid point answers for questions?

Answer 5

not in the persons knowledge/experience
meaningless questions
too personal/sensitive

Question 6

How do you avoid skewed distribution?

Answer 6

wording of the question:

I MIGHT find jury duty QUITE difficult; to

I WOULD find jury duty VERY difficult

Question 7

What is sentence mapping?

Answer 7

where you break a sentence down into facets and use the same core sentence with different facets to see how the different wording might affect answers.

e.g.

A COUNTRY has a RESOURCE which is USED IN A WAY for a CERTAIN PURPOSE

Question 8

What is facet theory

Answer 8

a framework for defining a research domain

Question 9

what is a prospective study

Answer 9

a study where you ask someone to self report their reactions to a situation

Question 10

what is a structuple?

Answer 10

a route through a mapping sentence

Question 11

Issues with structuples?

Answer 11

can lead to a high number of variables (>500 with 20 variables) so research must be limited with the possibility of other researchers working of different combinations of structuples

variables need to be interpreted consistently eg loved one could be different between wider family member or partner

variables need to be consistent and work together eg if you are with someone then further facets need to logically work, so cannot have items like ‘you were alone’

using terms like ‘loved one’ can be highly provocative

Question 12

Why do we avoid skewed distributions?

Answer 12

as this means the question is to (un)provoking

Question 13

What is the principle of contiguity?

Answer 13

more similar facets structures will be more similar empirically

Question 14

What is SSA?

Answer 14

Smallest space analysis - a form of multidimensional scaling
each point represents a questionnaire item
higher correlation, closer together the points are
shows overall relationships between questionnaire items
should be able to partition space for different elements