questionnaire design Flashcards

1
Q

Guidelines to constructing a questionnaire.

A
  1. make items clear
  2. avoid presuming questions
  3. simple language
  4. double barreled items
  5. respondents must be competent to answer
  6. avoid using not
  7. avoid using double negatives
  8. avoid using bias items
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
2
Q

0-100 scale

A

disadvantages

  • respondents answer in multiples, reducing the number of responses.
  • false precision (whats the difference between 36 and 38)
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
3
Q

visual analogue scale

A
advantage
very sensitive- 
difficult for respondents to remember previous responses
Dis
same as 0-100.
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
4
Q

worded response categories

A

dis

difficult to assess the psychological space between each word

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
5
Q

semantic differential scale

A

used to measure attitudes, using pairs of adjectives representing opposite ends of a continuum.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
6
Q

Thurstone scaling

A

Consists of items that have been designed to represent differing levels of an attribute.

  • large pool of items
  • judges sort them into piles
  • items in the piles scored 1 to 11 (median calculated)
  • 20 or so items are then selected.
  • items are presented in a random manner
  • attitude score is the average of the median values

Advantages
easy for respondent to complete
easy to develop alternative forms

Dis
labour intensive
judges ratings may differ from respondents

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
7
Q

guidelines for format

A
  1. white space
  2. general instructions
  3. question order
  4. introduce each section
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
8
Q

online survey tools pos and negs

A

pos

  • automatic
  • large samples
  • format into spss

neg

  • database of emails
  • low response rates
    • reduced power
    • selective/ biased sample
    • limits generalisability
How well did you know this?
1
Not at all
2
3
4
5
Perfectly