This document summarizes Pete Cape's lecture on questionnaire design. It discusses various types of questions that can be included in a questionnaire, such as open-ended, yes/no, single coded, and ranked questions. It also covers important principles for questionnaire design, such as remembering that respondents want to please the researcher and may be subject to biases like acquiescence bias. Finally, it addresses best practices for measuring concepts like frequencies and attitudes to obtain accurate responses.
VIP Call Girls In Green Park 9654467111 Escorts Service
Pete cape april lecture series - 2014
1. Pete Cape, SSI, UK
April Lecture Series 2014
Questionnaire Design – A Masterclass
Pete Cape
Global Knowledge Director, SSI
2014 Platinum Sponsor April Series Sponsor Lecture Sponsor
2. Pete Cape, SSI, UK
April Lecture Series 2014
| 2 |
A questionnaire might have all of these….
• Open questions
• Yes/No questions
• Single coded questions with multiple answers
• Multi-coded questions with multiple answers
• Items to be ranked
• Items to be rated
• Items to be chosen between
3. Pete Cape, SSI, UK
April Lecture Series 2014
| 3 |
A questionnaire might have all of these….
• Open questions
• Yes/No questions
• Single coded questions with multiple answers
• Multi-coded questions with multiple answers
• Items to be ranked
• Items to be rated
• Items to be chosen between
• But this is such a boring way of looking at it….
4. Pete Cape, SSI, UK
April Lecture Series 2014
| 4 |
A questionnaire might have all of these….
• Open questions
• Yes/No questions
• Single coded questions with multiple answers
• Multi-coded questions with multiple answers
• Items to be ranked
• Items to be rated
• Items to be chosen between
• But this is such a boring way of looking at it….
• Let’s start with some principles….
5. Pete Cape, SSI, UK
April Lecture Series 2014
| 5 |
Things to remember about respondents….
6. Pete Cape, SSI, UK
April Lecture Series 2014
| 6 |
Things to remember about respondents….
• They know they are in an experiment
7. Pete Cape, SSI, UK
April Lecture Series 2014
| 7 |
Things to remember about respondents….
• They know they are in an experiment
• Everything they see, hear or read is meant deliberately, and has
meaning
8. Pete Cape, SSI, UK
April Lecture Series 2014
| 8 |
Things to remember about respondents….
• They know they are in an experiment
• Everything they see, hear or read is meant deliberately, and has
meaning
• They want to please
9. Pete Cape, SSI, UK
April Lecture Series 2014
| 9 |
Things to remember about respondents….
• They know they are in an experiment
• Everything they see, hear or read is meant deliberately, and has
meaning
• They want to please
• They think you know what you are doing…
12. Pete Cape, SSI, UK
April Lecture Series 2014
| 12 |
Things to remember about questionnaires….
• Everything has meaning
13. Pete Cape, SSI, UK
April Lecture Series 2014
| 13 |
Things to remember about questionnaires….
• Everything has meaning
• You are supposed to know what that meaning is
14. Pete Cape, SSI, UK
April Lecture Series 2014
| 14 |
Questionnaires do a lot of work
• Motivate the respondent
15. Pete Cape, SSI, UK
April Lecture Series 2014
| 15 |
Questionnaires do a lot of work
• Motivate the respondent to provide
— complete answers
— accurate answers
16. Pete Cape, SSI, UK
April Lecture Series 2014
| 16 |
Questionnaires do a lot of work
• Motivate the respondent to provide
— complete answers
— accurate answers
• Communicate to the respondent
17. Pete Cape, SSI, UK
April Lecture Series 2014
| 17 |
Questionnaires do a lot of work
• Motivate the respondent to provide
— complete answers
— accurate answers
• Communicate to the respondent
• Help the respondent work out their answers
18. Pete Cape, SSI, UK
April Lecture Series 2014
| 18 |
Questionnaires do a lot of work
• Motivate the respondent to provide
— complete answers
— accurate answers
• Communicate to the respondent
• Help the respondent work out their answers
• Make the respondents task easy
19. Pete Cape, SSI, UK
April Lecture Series 2014
| 19 |
Is this the simplest question of all?
• Yes
• No
20. Pete Cape, SSI, UK
April Lecture Series 2014
| 20 |
Respondents want to please
Do you care to
disagree with
me….?
I am a scientist…..
• Acquiescence bias
21. Pete Cape, SSI, UK
April Lecture Series 2014
Acquiescence bias
• Some respondents are simply agreeable, and indicate
agreement out of politeness.
Saris, Krosnick and Shaeffe
22. Pete Cape, SSI, UK
April Lecture Series 2014
Acquiescence bias
• Some respondents are simply agreeable, and indicate
agreement out of politeness.
• Other respondents expect that the researchers agree with the
listed items and defer to their judgment.
Saris, Krosnick and Shaeffe
23. Pete Cape, SSI, UK
April Lecture Series 2014
Acquiescence bias
• Some respondents are simply agreeable, and indicate
agreement out of politeness.
• Other respondents expect that the researchers agree with the
listed items and defer to their judgment.
• Most respondents engage in survey satisficing and find that
agreeing takes less effort than carefully weighing each
optional level of disagreement and agreement.
Saris, Krosnick and Shaeffe
24. Pete Cape, SSI, UK
April Lecture Series 2014
Affirmation vs Multicode
25. Pete Cape, SSI, UK
April Lecture Series 2014
Affirmation vs Multicode
26. Pete Cape, SSI, UK
April Lecture Series 2014
Overcoming acquiescence bias
• Find them through traps and exclude
27. Pete Cape, SSI, UK
April Lecture Series 2014
Overcoming acquiescence bias
• Find them through traps and exclude
• Follow up question “how sure?”
• If not 100% sure then recode as “no”
28. Pete Cape, SSI, UK
April Lecture Series 2014
Overcoming acquiescence bias
• Find them through traps and exclude
• Follow up question “how sure?”
• If not 100% sure then recode as “no”
• Avoid the question type and give alternatives
— “chose the one of these that best describes you”
— “to what extent do you…”
29. Pete Cape, SSI, UK
April Lecture Series 2014
Affirmation vs Multicode
30. Pete Cape, SSI, UK
April Lecture Series 2014
Affirmation vs Multicode
why so big…?
31. Pete Cape, SSI, UK
April Lecture Series 2014
Social Desirability bias
• Wanting to look good
32. Pete Cape, SSI, UK
April Lecture Series 2014
Social Desirability bias
• Wanting to look good
— To society
— To other individuals
— To myself
33. Pete Cape, SSI, UK
April Lecture Series 2014
Some Normal Everyday Things…that I keep to myself..
Chart Title
34. Pete Cape, SSI, UK
April Lecture Series 2014
Some Normal Everyday Things…I want you to think I do..
Chart Title
35. Pete Cape, SSI, UK
April Lecture Series 2014
20% bought a book last week?
| 35 |
36. Pete Cape, SSI, UK
April Lecture Series 2014
Avoiding Social Desirability
• Is very hard
• Try to make respondents feel “normal”
37. Pete Cape, SSI, UK
April Lecture Series 2014
Avoiding Social Desirability
• Is very hard
• Try to make respondents feel “normal”
This third section is about social attitudes and values. There is a
lot of talk about discrimination and racial prejudice, and different
people have different views on the size of such problems, their
causes and what should be done about them. There are no right
or wrong answers, it is your honest opinions we are looking for. If
you do not wish to answer any of these questions please use the
option “prefer not to say”.
38. Pete Cape, SSI, UK
April Lecture Series 2014
Questionnaire Design and Bias
• Questionnaire design has to overcome respondents’
psychological biases
39. Pete Cape, SSI, UK
April Lecture Series 2014
Questionnaire Design and Bias
• Questionnaire design has to overcome respondents’
psychological biases
• And also experimenter bias
40. Pete Cape, SSI, UK
April Lecture Series 2014
Questionnaire Design and Bias
• Questionnaire design has to overcome respondents’
psychological biases
• And also experimenter bias
• i.e. you and your client….
41. Pete Cape, SSI, UK
April Lecture Series 2014
Experimenter biases
• It is easy to suggest the answer…
42. Pete Cape, SSI, UK
April Lecture Series 2014
Question suggests the answer
Which carbonated soft drinks, like Coca-Cola or Pepsi
for example, do you drink regularly?”
43. Pete Cape, SSI, UK
April Lecture Series 2014
Question suggests the answer
Which carbonated soft drinks, like Coca-Cola or Pepsi
for example, do you drink regularly?”
Raises answers for drinks
“like” Pepsi and Coke
i.e. Brown or cola drinks
44. Pete Cape, SSI, UK
April Lecture Series 2014
Question suggests the answer
Which carbonated soft drinks, like Coca-Cola or Pepsi
for example, do you drink regularly?”
Raises answers for drinks
“like” Pepsi and Coke
i.e. Brown or cola drinks
Lowers answers for non-
brown drinks
e.g. Fanta, 7-Up
45. Pete Cape, SSI, UK
April Lecture Series 2014
Question suggests the answer
Which carbonated soft drinks, like Coca-Cola or Pepsi
for example, do you drink regularly?”
Raises answers for drinks
“like” Pepsi and Coke
i.e. Brown or cola drinks
Lowers answers for non-
brown drinks
e.g. Fanta, 7-Up
Problem is in use of jargon term: “carbonated soft drinks”
46. Pete Cape, SSI, UK
April Lecture Series 2014
Question suggests “my” answer
Testing perfumes on animals involves causing them
pain. Do you agree with testing perfumes on animals?
47. Pete Cape, SSI, UK
April Lecture Series 2014
Question suggests “my” answer
Testing perfumes on animals involves causing them
pain. Do you agree with testing perfumes on animals?
“involves causing them pain” is an opinion
48. Pete Cape, SSI, UK
April Lecture Series 2014
Avoiding Experimenter Bias
• Is easier
49. Pete Cape, SSI, UK
April Lecture Series 2014
Avoiding Experimenter Bias
• Is easier
• Think neutral
50. Pete Cape, SSI, UK
April Lecture Series 2014
Avoiding Experimenter Bias
• Is easier
• Think neutral
• Don’t think your view is the world view
51. Pete Cape, SSI, UK
April Lecture Series 2014
Questionnaire as a measuring tool
• We rarely measure physical things
52. Pete Cape, SSI, UK
April Lecture Series 2014
Questionnaire as a measuring tool
• We rarely measure physical things
• But we do measure frequency:
— How often do you x, y, z?
53. Pete Cape, SSI, UK
April Lecture Series 2014
Questionnaire as a measuring tool
• We rarely measure physical things
• But we do measure frequency:
— How often do you x, y, z?
• And we do measure attitudes:
— How much do you think x,y,z?
54. Pete Cape, SSI, UK
April Lecture Series 2014
Measuring frequency
• As an relative concept in a fixed time period
55. Pete Cape, SSI, UK
April Lecture Series 2014
Measuring frequency
• As a relative concept in a fixed time period
• Define:
— The activity
• What does it mean to “go shopping”?
• What does it mean to “surf the internet?”
— The time period
• Is it the relevant one?
• What is an “average week”?
— The measurement scale
• Always, often, sometimes, never…..
56. Pete Cape, SSI, UK
April Lecture Series 2014
Measuring frequency
• As an absolute number in a fixed time period
57. Pete Cape, SSI, UK
April Lecture Series 2014
Measuring frequency
• As an absolute number in a fixed time period
• Define:
— The activity
• What does it mean to “go shopping”?
• What does it mean to “surf the internet?”
— The time period
• Is it the relevant one?
• What is an “average week”?
58. Pete Cape, SSI, UK
April Lecture Series 2014
Measuring frequency
• As an absolute number in a fixed time period
How many times did you take a bath in the past year?
59. Pete Cape, SSI, UK
April Lecture Series 2014
Measuring frequency
• As an absolute number in a fixed time period
How many times did you take a bath in the past year?
too longwhat does
this mean?
60. Pete Cape, SSI, UK
April Lecture Series 2014
Measuring frequency
• Precision and correct framing
How many times, if at all, did you take a bath in the past week? By
taking a bath we mean actually using a bath tub and filling it with
water and sitting in it. Please exclude any showers you may have taken
61. Pete Cape, SSI, UK
April Lecture Series 2014
Measuring frequency
• Precision and correct framing
How many times, if at all, did you take a bath in the past week? By
taking a bath we mean actually using a bath tub and filling it with
water and sitting in it. Please exclude any showers you may have taken
OMG, I only take showers,
he will think I am so dirty
62. Pete Cape, SSI, UK
April Lecture Series 2014
Measuring frequency
• Precision and correct framing, and overcoming social desirability bias
Some people only take baths, some only take showers, some take both.
This part of the survey is only concerned with bath taking.
How many times, if at all, did you take a bath in the past week? By
taking a bath we mean actually using a bath tub and filling it with
water and sitting in it. Please exclude any showers you may have taken
63. Pete Cape, SSI, UK
April Lecture Series 2014
Measuring frequency
• Precision and correct framing, and overcoming social desirability bias, and to
make life easy
Some people only take baths, some only take showers, some take both.
This part of the survey is only concerned with bath taking.
How many times, if at all, did you take a bath in the past week? By
taking a bath we mean actually using a bath tub and filling it with
water and sitting in it. Please exclude any showers you may have taken
0 / only took showers
1 – 2
3 – 4
5 – 6
7 or more
Can’t remember
64. Pete Cape, SSI, UK
April Lecture Series 2014
Measuring frequency
• Precision and correct framing, and overcoming social desirability bias, and to
make life easy
Some people only take baths, some only take showers, some take both.
This part of the survey is only concerned with bath taking.
How many times, if at all, did you take a bath in the past week? By
taking a bath we mean actually using a bath tub and filling it with
water and sitting in it. Please exclude any showers you may have taken
0 / only took showers
1 – 2
3 – 4
5 – 6
7 or more
Can’t remember
Does this have any meaning?
65. Pete Cape, SSI, UK
April Lecture Series 2014
The meaning of ranges
66. Pete Cape, SSI, UK
April Lecture Series 2014
The meaning of ranges
• Assumption of normal distribution
67. Pete Cape, SSI, UK
April Lecture Series 2014
The meaning of ranges
• Assumption of normal distribution
68. Pete Cape, SSI, UK
April Lecture Series 2014
Manipulating numbers
Answer list A Answer list B
69. Pete Cape, SSI, UK
April Lecture Series 2014
Manipulating numbers
Answer list A Answer list B
70. Pete Cape, SSI, UK
April Lecture Series 2014
Manipulating numbers
Answer list A Answer list B
71. Pete Cape, SSI, UK
April Lecture Series 2014
Manipulating numbers
Answer list A
• The central banding should contain the mean/median
• But we don’t know it, it’s what we’re trying to find out!
Answer list B
72. Pete Cape, SSI, UK
April Lecture Series 2014
Collecting real numbers
• Just ask for it
• Respondent knows it and will give it
73. Pete Cape, SSI, UK
April Lecture Series 2014
Measuring attitudes
74. Pete Cape, SSI, UK
April Lecture Series 2014
Measuring attitudes
Fill in the words that correspond to the scale points for “Likely”
0 = not at all likely
1 = ?
2 = ?
3 = ?
4 = ?
5 = ?
6 = ?
7 = ?
8 = ?
9 = ?
10 = extremely likely
?
?
Exercise
75. Pete Cape, SSI, UK
April Lecture Series 2014
Measuring attitudes
Fill in the words that correspond to the scale points for “Likely”
0 = not at all likely
1 = ?
2 = ?
3 = ?
4 = ?
5 = ?
6 = ?
7 = ?
8 = ?
9 = ?
10 = extremely likely
Exercise
76. Pete Cape, SSI, UK
April Lecture Series 2014
Measuring attitudes
Fill in the words that correspond to the scale points for “Likely”
0 = not at all likely
1 = ?
2 = ?
3 = ?
4 = ?
5 = ?
6 = ?
7 = ?
8 = ?
9 = ?
10 = extremely likely
Exercise
Exercise outcomes:
a) No-one will be able to do it
b) Some people will write:
5 = “neither likely not unlikely”
5 = “some degree of likelihood”
77. Pete Cape, SSI, UK
April Lecture Series 2014
Measuring attitudes
Fill in the words that correspond to the scale points for “Likely”
0 = not at all likely
1 = ?
2 = ?
3 = ?
4 = ?
5 = ?
6 = ?
7 = ?
8 = ?
9 = ?
10 = extremely likely
Exercise
Exercise outcomes:
a) No-one will be able to do it
b) Some people will write:
5 = “neither likely nor unlikely”
5 = “some degree of likelihood”
78. Pete Cape, SSI, UK
April Lecture Series 2014
Two men go into a bar….
79. Pete Cape, SSI, UK
April Lecture Series 2014
Two men go into a bar….
I say, what do
you think of
the new VW
Golf?
80. Pete Cape, SSI, UK
April Lecture Series 2014
Two men go into a bar….
I say, what do
you think of
the new VW
Golf?
I think
it’s a 6…
81. Pete Cape, SSI, UK
April Lecture Series 2014
Two men go into a bar….
Attitudes are expressed
in words not numbers
I say, what do
you think of
the new VW
Golf?
I think
it’s a 6…
82. Pete Cape, SSI, UK
April Lecture Series 2014
Numbers are universal
83. Pete Cape, SSI, UK
April Lecture Series 2014
Numbers are universal
• Numeric scales are not!
84. Pete Cape, SSI, UK
April Lecture Series 2014
Numbers are universal
• Numeric scales are not!
• It all starts in school
— 0 = you weren’t even there!
— 1-4 = you might as well not have been there…
— 5-6 = you are going to have to work way harder in the future
— 7-8 = okay, could do better
— 9 = good
— 10 = very good
85. Pete Cape, SSI, UK
April Lecture Series 2014
Numbers are universal
• Numeric scales are not!
• It all starts in school
— 0 = you weren’t even there!
— 1-4 = you might as well not have been there…
— 5-6 = you are going to have to work way harder in the future
— 7-8 = okay, could do better
— 9 = good
— 10 = very good
• And this is for cultures that score 0 - 10, does everyone?
86. Pete Cape, SSI, UK
April Lecture Series 2014
German School System
• 1 (sehr gut, very good) is the best possible grade and is given for
outstanding performance
• 2 (gut, good) is the next-highest and is given for performance that
meets the standard completely and is above-average
• 3 (befriedigend, satisfactory) indicates "average" performance.
• 4 (ausreichend, sufficient) is the lowest passing grade and is given if
the standard has been met but with a number of notable errors.
• 5 (mangelhaft, deficient) is the higher of two failing grades and is
given if the standard has not been met but the basics have been
understood.
• 6 (ungenügend, insufficient) is the lowest possible grade and is given
if the standard has not been met and the basics have not been
understood.
87. Pete Cape, SSI, UK
April Lecture Series 2014
Numbers have meaning
0 1 2 3 4 5 6 7 8 9 10
Schwartz, 1996
88. Pete Cape, SSI, UK
April Lecture Series 2014
Numbers have meaning
0 1 2 3 4 5 6 7 8 9 10
34%
Schwartz, 1996
89. Pete Cape, SSI, UK
April Lecture Series 2014
Numbers have meaning
0 1 2 3 4 5 6 7 8 9 10
-5 -4 -3 -2 -1 0 1 2 3 4 5
34%
Schwartz, 1996
90. Pete Cape, SSI, UK
April Lecture Series 2014
Numbers have meaning
0 1 2 3 4 5 6 7 8 9 10
-5 -4 -3 -2 -1 0 1 2 3 4 5
34%
13%?
Schwartz, 1996
91. Pete Cape, SSI, UK
April Lecture Series 2014
Numbers have meaning
0 1 2 3 4 5 6 7 8 9 10
-5 -4 -3 -2 -1 0 1 2 3 4 5
34%
13%
Schwartz, 1996
92. Pete Cape, SSI, UK
April Lecture Series 2014
This what scales are trying to achieve
93. Pete Cape, SSI, UK
April Lecture Series 2014
This what scales are trying to achieve
94. Pete Cape, SSI, UK
April Lecture Series 2014
• Turn to academia for the answer
How long should a scale be?
95. Pete Cape, SSI, UK
April Lecture Series 2014
• Turn to academia for the answer
• Is not always shared
• But is somewhere between 5 and 9
points
How long should a scale be?
96. Pete Cape, SSI, UK
April Lecture Series 2014
• Turn to academia for the answer
• Is not always shared
• But is somewhere between 5 and 9
points
• Jon Krosnick says:
• Unipolar: 5 points
• Bipolar: 7 points
How long should a scale be?
97. Pete Cape, SSI, UK
April Lecture Series 2014
• Turn to academia for the answer
• Is not always shared
• But is somewhere between 5 and 9
points
• Jon Krosnick says:
• Unipolar: 5 points
• Bipolar: 7 points
• Why?
• Maximises Reliability and Validity
How long should a scale be?
98. Pete Cape, SSI, UK
April Lecture Series 2014
What is reliability and validity?
• Reliability
— Measures the same thing in the same way every time
99. Pete Cape, SSI, UK
April Lecture Series 2014
What is reliability and validity?
• Reliability
— Measures the same thing in the same way every time
• Validity
— Measures what you think it measures
100. Pete Cape, SSI, UK
April Lecture Series 2014
What was Unipolar and Bipolar?
“Unipolar”
One direction from zero to lots
he middle and surfeit at the ends
101. Pete Cape, SSI, UK
April Lecture Series 2014
What was Unipolar and Bipolar?
“Unipolar”
One direction from zero to lots
“Bipolar”
Opposites with indifference in the middle
he middle and surfeit at the ends
102. Pete Cape, SSI, UK
April Lecture Series 2014
Scales
Unipolar:
• Extremely x
• Very x
• Somewhat x
• Slightly x
• Not at all x
(an) academic conclusion
103. Pete Cape, SSI, UK
April Lecture Series 2014
Scales
Unipolar:
• Extremely x
• Very x
• Somewhat x
• Slightly x
• Not at all x
Bipolar:
• Extremely x
• Very x
• Somewhat x
• Neither x nor un-x
• Somewhat un-x
• Very un-x
• Extremely un-x
(an) academic conclusion
104. Pete Cape, SSI, UK
April Lecture Series 2014
Scales
Unipolar:
• Extremely x
• Very x
• Somewhat x
• Slightly x
• Not at all x
Bipolar:
• Extremely x
• Very x
• Somewhat x
• Neither x nor un-x
• Somewhat un-x
• Very un-x
• Extremely un-x
And this is what I recommend when asked....
btw: you can’t mix bi- and unipolar scales
(an) academic conclusion
105. Pete Cape, SSI, UK
April Lecture Series 2014
Scales
• Design issues
• Display issues
106. Pete Cape, SSI, UK
April Lecture Series 2014
Where to put “don’t know”
• Visual and Conceptual midpoint
107. Pete Cape, SSI, UK
April Lecture Series 2014
Where to put “don’t know”
• Visual and Conceptual midpoint
Conceptual midpoint
Visual
midpoint
108. Pete Cape, SSI, UK
April Lecture Series 2014
Where to put “don’t know”
• Visual and Conceptual midpoint
109. Pete Cape, SSI, UK
April Lecture Series 2014
Where to put “don’t know”
• Visual and Conceptual midpoint
Visual
midpoint
110. Pete Cape, SSI, UK
April Lecture Series 2014
Where to put “don’t know”
• Visual and Conceptual midpoint
• Skews data towards “disagree”
Conceptual midpoint
Visual
midpoint
111. Pete Cape, SSI, UK
April Lecture Series 2014
Where to put “don’t know”
• Visual and Conceptual midpoint
Conceptual midpoint Visual
midpoint
112. Pete Cape, SSI, UK
April Lecture Series 2014
What order to present?
Disagree
Strongly
Disagree
Slightly
Neither Agree
nor Disagree
Agree
Slightly
Agree
Strongly
113. Pete Cape, SSI, UK
April Lecture Series 2014
What order to present?
Disagree
Strongly
Disagree
Slightly
Neither Agree
nor Disagree
Agree
Slightly
Agree
Strongly
Order bias
114. Pete Cape, SSI, UK
April Lecture Series 2014
What order to present?
Disagree
Strongly
Disagree
Slightly
Neither Agree
nor Disagree
Agree
Slightly
Agree
Strongly
Order bias Acquiescence bias
115. Pete Cape, SSI, UK
April Lecture Series 2014
What order to present?
Disagree
Strongly
Disagree
Slightly
Neither Agree
nor Disagree
Agree
Slightly
Agree
Strongly
Order bias Acquiescence bias
Central tendency
116. Pete Cape, SSI, UK
April Lecture Series 2014
What order to present?
Disagree
Strongly
Disagree
Slightly
Neither Agree
nor Disagree
Agree
Slightly
Agree
Strongly
Order bias Acquiescence bias
Central tendency
Agree
Strongly Agree Slightly
Neither Agree
nor Disagree
Disagree
Slightly
Disagree
Strongly
Central tendency
Order bias
Acquiescence bias
117. Pete Cape, SSI, UK
April Lecture Series 2014
Where to put the answer box
• Next to answers
• 22% failure rate on matching gender to panel record!
118. Pete Cape, SSI, UK
April Lecture Series 2014
Where to put the answer box
• Next to answers
• 22% failure rate on matching gender to panel record!
• Left right processing
• 98% of that: Males coded as Females
119. Pete Cape, SSI, UK
April Lecture Series 2014
Scales
• Maximises reliability and validity
• Minimises bias
• Prevents order error
• Not just “the way we always do it…”
120. Pete Cape, SSI, UK
April Lecture Series 2014
Scales
• Maximises reliability and validity
• Minimises bias
• Prevents order error
• Not just “the way we always do it…”
• But isn’t it old-fashioned?
121. Pete Cape, SSI, UK
April Lecture Series 2014
Sliders with visual elements
122. Pete Cape, SSI, UK
April Lecture Series 2014
129125
77
25
4
0
50
100
150
200
250
Different styles of slider scale
123. Pete Cape, SSI, UK
April Lecture Series 2014
Different styles of slider scale
114
8892
17
1
0
50
100
150
200
250
124. Pete Cape, SSI, UK
April Lecture Series 2014
Different styles of slider scale
97
21
27
8
1
0
50
100
150
200
250
125. Pete Cape, SSI, UK
April Lecture Series 2014
4
10
124
4751
0
50
100
150
200
250
Different styles of slider scale
126. Pete Cape, SSI, UK
April Lecture Series 2014
Different distributions, same means?
UK USA Germany China
traditional - raw data 2.27 2.32 2.50 3.79
traditional - rescored 2.60 2.28 2.50 3.87
slider with all labels 2.56 2.64 2.74 3.89
slider with all tick marks 2.33 2.35 2.46 3.61
slider with no tick marks 2.32 2.39 2.29 3.62
slider with no tick marks, score showing 2.34 2.44 2.33 3.56
127. Pete Cape, SSI, UK
April Lecture Series 2014
Grids and Matrices
• Q: Do we have these in telephone research?
• Q: In Face to Face?
128. Pete Cape, SSI, UK
April Lecture Series 2014
Grids and Matrices
• Q: Do we have these in telephone research?
• Q: In Face to Face?
• A: No
• Q: Where did we have them?
129. Pete Cape, SSI, UK
April Lecture Series 2014
Grids and Matrices
• Q: Do we have these in telephone research?
• Q: In Face to Face?
• A: No
• Q: Where did we have them?
• A: postal research
130. Pete Cape, SSI, UK
April Lecture Series 2014
Item-in-a-series
131. Pete Cape, SSI, UK
April Lecture Series 2014
Item-in-a-series
“When one has to ask a series of
questions that use the same answer
categories it is convenient to combine
them into a [grid/matrix format].
Combining them into this ‘item-in-a-
series’, with a common introduction
that defines the general question and
response format, eliminates
considerable redundancy with regard
to stating questions, it also saves
considerable space”
132. Pete Cape, SSI, UK
April Lecture Series 2014
Item-in-a-series
“but do it carefully”
133. Pete Cape, SSI, UK
April Lecture Series 2014
Item-in-a-series
“but do it carefully”
• Items are now in a comparative
framework
• Visual rendering makes them a unit
“if the sponsor wants individuals to
contemplate each item separately, it is
advisable to present each of them as
individual items”
134. Pete Cape, SSI, UK
April Lecture Series 2014
Item-in-a-series
Should these should be the same?
135. Pete Cape, SSI, UK
April Lecture Series 2014
Item-in-a-series
Shouldn’t these have been the same?
136. Pete Cape, SSI, UK
April Lecture Series 2014
| 136 |
What else do grids do?
• Encourage you to go too fast
137. Pete Cape, SSI, UK
April Lecture Series 2014
Speed of grid completion
•What sort of processing is being done at this speed?
| 137 |
138. Pete Cape, SSI, UK
April Lecture Series 2014
| 138 |
What else do grids do?
• Encourage you to go too fast
• Process the underlying latent construct, not the items
— Straightlining
139. Pete Cape, SSI, UK
April Lecture Series 2014
| 139 |
What else do grids do?
• Encourage you to go too fast
• Process the underlying latent construct, not the items
— Straightlining
• Miss the subtleties
— Heuristic answering
140. Pete Cape, SSI, UK
April Lecture Series 2014
Direct instruction at item 39/40 – pass/fail
| 140 |
141. Pete Cape, SSI, UK
April Lecture Series 2014
Which heuristic was used?
| 141 |
answer at trap
142. Pete Cape, SSI, UK
April Lecture Series 2014
Which heuristic was used?
| 142 |
answer at trap
143. Pete Cape, SSI, UK
April Lecture Series 2014
And those left over?
| 143 |
144. Pete Cape, SSI, UK
April Lecture Series 2014
And those left over?
| 144 |
145. Pete Cape, SSI, UK
April Lecture Series 2014
| 145 |
Anything else wrong with grids?
• Yes, we use the wrong scale for the construct
146. Pete Cape, SSI, UK
April Lecture Series 2014
Beware of the agree-disagree question
• Make your scale answers the appropriate ones
To what extent to you agree or disagree that this
training course is extremely useful?
Agree extremely
Very much agree
Somewhat agree
Neither agree nor disagree
Somewhat disagree
Very much disagree
Disagree extremely
147. Pete Cape, SSI, UK
April Lecture Series 2014
2 stage process
1. Decide how useful you find this
148. Pete Cape, SSI, UK
April Lecture Series 2014
2 stage process
1. Decide how useful you find this
2. Translate into agree-disagree, hopefully:
Extremely useful = Agree extremely
Very useful = Very much agree
Somewhat useful = Somewhat agree
Neither useful not disuseful = Neither agree nor disagree
Somewhat disuseful = Somewhat disagree
Very disuseful = Very much disagree
Extremely disuseful = Disagree extremely
149. Pete Cape, SSI, UK
April Lecture Series 2014
2 stage process
1. Decide how useful you find this
2. Translate into agree-disagree, hopefully:
• Disuseful?
Extremely useful = Agree extremely
Very useful = Very much agree
Somewhat useful = Somewhat agree
Neither useful not disuseful = Neither agree nor disagree
Somewhat disuseful = Somewhat disagree
Very disuseful = Very much disagree
Extremely disuseful = Disagree extremely
150. Pete Cape, SSI, UK
April Lecture Series 2014
2 stage process
• Reality (maybe):
Extremely useful = Agree extremely
Very useful = Very much agree
Somewhat useful = Somewhat agree
Slightly useful = Somewhat agree
Not at all useful = Neither/nor, Somewhat disagree,Very
much disagree,Disagree extremely
151. Pete Cape, SSI, UK
April Lecture Series 2014
How to measure attitudes/constructs
• Work out the underlying construct
— i.e. what are you measuring
152. Pete Cape, SSI, UK
April Lecture Series 2014
How to measure attitudes/constructs
• Work out the underlying construct
— i.e. what are you measuring
• Decide how it is measured
— What words
— Unipolar or Bipolar
153. Pete Cape, SSI, UK
April Lecture Series 2014
How to measure attitudes/constructs
• Work out the underlying construct
— i.e. what are you measuring
• Decide how it is measured
— What words
— Unipolar or Bipolar
• Check your measurement gaps are equal
• Ask question..
154. Pete Cape, SSI, UK
April Lecture Series 2014
How to measure attitudes/constructs
• Work out the underlying construct
— i.e. what are you measuring
• Decide how it is measured
— What words
— Unipolar or Bipolar
• Check your measurement gaps are equal
• Ask question..
• More work for you, less work for respondent
• And more valid and reliable
155. Pete Cape, SSI, UK
April Lecture Series 2014
Construct Specific Scale
How useful, if at all, did you find this training course?
Extremely useful
Very useful
Somewhat useful
Slightly useful
Not at all useful
• As you read the question your answer is forming…..
?
156. Pete Cape, SSI, UK
April Lecture Series 2014
Construct Specific Scale
How useful, if at all, did you find this training course?
Extremely useful
Very useful
Somewhat useful
Slightly useful
Not at all useful
• As you read the question your answer is forming…..
• Order effects?
157. Pete Cape, SSI, UK
April Lecture Series 2014
Construct Specific Scale
Extremely useful
Very useful
Somewhat useful
Slightly useful
Not at all useful
158. Pete Cape, SSI, UK
April Lecture Series 2014
Construct Specific Scale
Extremely useful
Very useful
Somewhat useful
Slightly useful
Not at all useful
Expectation: up = good
(Tourangeau)
159. Pete Cape, SSI, UK
April Lecture Series 2014
Construct Specific Scale
Extremely useful
Very useful
Somewhat useful
Slightly useful
Not at all useful
Expectation: up = good
(Tourangeau)
Answer speed = faster
(Christian et al)
No order bias found
160. Pete Cape, SSI, UK
April Lecture Series 2014
Manipulating people by accident
• Because we want to make it more fun
161. Pete Cape, SSI, UK
April Lecture Series 2014
Please select your most preferred holiday type…
Summer Beach Winter
City Break Adventure
162. Pete Cape, SSI, UK
April Lecture Series 2014
Please select your most preferred holiday type…
Summer Beach Winter
City Break Adventure
163. Pete Cape, SSI, UK
April Lecture Series 2014
Please select your most preferred holiday type…
Summer Beach Winter
City Break Adventure
164. Pete Cape, SSI, UK
April Lecture Series 2014
The results
58%
60%
36%
0%
20%
40%
60%
80%
100%
“Summer Beach”
166. Pete Cape, SSI, UK
April Lecture Series 2014
Motivation
Self-Determination Theory (Deci & Ryan)
167. Pete Cape, SSI, UK
April Lecture Series 2014
Motivation
• people are active ‘organisms’
• having tendencies toward psychological growth
and development
• who strive to master ongoing challenges
• and to integrate their experiences into a
coherent sense of self.
• the social context can either support or thwart
the natural tendencies
Self-Determination Theory (Deci & Ryan)
168. Pete Cape, SSI, UK
April Lecture Series 2014
Motivation under Self-Determination Theory
A continuum that relates to the task in
the social context, not a person
169. Pete Cape, SSI, UK
April Lecture Series 2014
Motivation under Self-Determination Theory
A continuum that relates to the task in
the social context, not a person
Regulation
Introjection
Identification
Integration
to obtain an externally imposed reward
pressure to avoid guilt or to enhance self-esteem
conscious valuing of activity
assimilated to self
task is not done
for it’s own reward
170. Pete Cape, SSI, UK
April Lecture Series 2014
Motivation under Self-Determination Theory
A continuum that relates to the task in
the social context, not a person
Regulation
Introjection
Identification
Integration
to obtain an externally imposed reward
pressure to avoid guilt or to enhance self-esteem
conscious valuing of activity
assimilated to self
task is not done
for it’s own reward
Where do most market research
respondents fall?
171. Pete Cape, SSI, UK
April Lecture Series 2014
Motivation under Self-Determination Theory
A continuum that relates to the task in
the social context, not a person
Regulation
Introjection
Identification
Integration
to obtain an externally imposed reward
pressure to avoid guilt or to enhance self-esteem
conscious valuing of activity
assimilated to self
task is not done
for it’s own reward
Where do most market research
respondents fall?
172. Pete Cape, SSI, UK
April Lecture Series 2014
Motivation under Self-Determination Theory
Task Outcomes
Regulation
Introjection
Identification
Integration
little interest, value or effort
more effort, anxiety, poor coping with failure
enjoyment, good coping skills
greater enjoyment, competency
task is not done properly
interest, enjoyment
highly competent
Where would we like respondents
to be?
173. Pete Cape, SSI, UK
April Lecture Series 2014
Motivation under Self-Determination Theory
Task Outcomes
Regulation
Introjection
Identification
Integration
little interest, value or effort
more effort, anxiety, poor coping with failure
enjoyment, good coping skills
greater enjoyment, competency
task is not done properly
interest, enjoyment
highly competent
Where would we like respondents
to be?
174. Pete Cape, SSI, UK
April Lecture Series 2014
Moving people along the motivation continuum
• Foster feelings of:
— Autonomy – you are free to do this or not as you choose
— Competence – you are good at this
— Relatedness – people like you do this
— Value – what you are doing has meaning
175. Pete Cape, SSI, UK
April Lecture Series 2014
Moving people along the motivation continuum
• Foster feelings of:
— Autonomy – you are free to do this or not as you choose
— Competence – you are good at this
— Relatedness – people like you do this
— Value – what you are doing has meaning
• Perhaps this was what the interviewer used to do?
• This is what Gamification helps do
176. Pete Cape, SSI, UK
April Lecture Series 2014
Motivating respondents
Thank you very much for choosing to do this important survey.
As one of our most expert survey takers we have selected you to
help on this survey.
People like you all over the country are taking part in this survey.
Your answers and those of everyone else will help improve the
services our client provides to customers like you.
Please press the > button when you are ready to continue with
your survey.
177. Pete Cape, SSI, UK
April Lecture Series 2014
Motivating respondents
Thank you very much for choosing to do this important survey.
As one of our most expert survey takers we have selected you to
help on this survey.
People like you all over the country are taking part in this survey.
Your answers and those of everyone else will help improve the
services our client provides to customers like you.
Please press the > button when you are ready to continue with
your survey.
• No mention of rewards
• No threats
178. Pete Cape, SSI, UK
April Lecture Series 2014
Summary
• The questionnaire is a mixture of art and science
• The art of writing well, of motivating respondents
• The science of writing precisely, avoiding bias
• Everything has meaning
• You are supposed to know what it means
• There are better ways of asking questions
• There are better ways of collecting answers
• You might as well do the best you can
179. Pete Cape, SSI, UK
April Lecture Series 2014
Summary
• Questionnaire writing
• (and understanding why it is important)
• Sets us apart from others
• And has real value
180. Pete Cape, SSI, UK
April Lecture Series 2014
Summary - Principles
• Respondents know they are in an experiment
• They think everything they see, hear or read is deliberate, and
has meaning
• They want to please
• Questionnaire motivates the respondent to provide
— complete answers
— accurate answers
• It communicates to the respondent
• It helps the respondent work out their answers
• It makes the respondents task easy
181. Pete Cape, SSI, UK
April Lecture Series 2014
And guess what?
• If it was easy do you think there
would be textbooks on it?
182. Pete Cape, SSI, UK
April Lecture Series 2014
Thank you!
Pete Cape
Global Knowledge Director, SSI
2014 Platinum Sponsor April Series Sponsor Lecture Sponsor
183. Pete Cape, SSI, UK
April Lecture Series 2014
The Sponsors for this Event
If you are interested in sponsoring a future NewMR event
Email Michele.Poynter@TheFuturePlace.com
2014 Platinum Sponsor April Series Sponsor Lecture Sponsor
184. Pete Cape, SSI, UK
April Lecture Series 2014
Q & A
Ray Poynter
The Future Place
Pete Cape
SSI
2014 Platinum Sponsor April Series Sponsor Lecture Sponsor
185. Pete Cape, SSI, UK
April Lecture Series 2014
| 186 |
surveysampling.com
info@surveysampling.com
pete.cape@surveysampling.com