You are on page 1of 22

M K T 450 MA RK ET IN G T O O L S

Buyer Behavior
and Market Analysis

Questionnaire Development

M KT 450 Introductory Remarks

Questionnaires collect data two ways:

1. Open-ended questioning (unstructured, unlimited


response options)

2. Closed-ended questioning (structured, limited,


predetermined response options )

1
M KT 450 Introductory Remarks

Our focus will be on closed ended-questioning.

Closed-ended questionnaire items are composed of

questions: portion of the item that asks for or instructs


respondent to give particular information

scales: provides respondents with possible responses


to questions

M KT 450 Introductory Remarks

Example
:
Very Very
Dissatisfie Satisfie
Rate your degree of overall
d d
satisfaction with your meal.
-2 0 +1
+2
-1
question scale

ite
m

2
M KT 450 Introductory Remarks

Example
:
Strongly Strongly
Agree
I enjoy trying new types of
Disagree
food. 3 4
1 5

2
question scale

ite
m

M KT 450 Introductory Remarks

Example
:

Are you. . . Male? Female?

question scale

ite
m

3
M KT 450 Introductory Remarks

Basic Types of Questionnaire Items

category items

semantic differential items

Likert items

constant sum items

yes/no items

checklist items

M KT 450 Item Development Principles

1. Avoid confusing items.


Speak to your respondents in their language in as simple terms
as possible.

Confusing items can be incomprehensible.


Given current economic conditions to what extent have you
or members of your household altered ordinary
discretionary consumption patterns, excepting necessities
other than staples?
Not at Very A
Great
All Little
Deal
0 1 2 3 4 5

4
M KT 450 Item Development Principles

1. Avoid confusing items.


Speak to your respondents in their language in as simple terms
as possible.

Confusing items can be incomprehensible.

Confusing items can be unanswerable.


Over the previous year, what has been the average wait
time for seating at local restaurants?
No Very Very
Wait Short
Long
0 1 2 3 4 5

M KT 450 Item Development Principles

2. Avoid oversimplification.

Be as specific as possible to avoid multiple interpretations.


Strongly Strongly
Agree
Disagree
I enjoy going to concerts. 1 2 3 4 5
I frequently go to clubs. 1 2 3 4 5
I’m good with computers. 1 2 3 4 5

5
M KT 450 Item Development Principles

3. Generally speaking, avoid “double-barreled” items.

Each item should refer to a single, specific idea.

How important are quality and reputation in selecting clothes


for your children?
Not At All Slightly Very
Important Important Important
0 1 2 3 4 5

M KT 450 Item Development Principles

4. Avoid leading items.

Items should be as objective and value-neutral as possible.

Does knowing that Speedplay is the highest rated extreme


sports equipment make you more confident in your purchase?
Yes
No

The tax and spend liberals in Congress will lead our country
to economic ruin.
Strongly Strongly
1 2 3 4 5
Disagree
Agree

6
M KT 450 Item Development Principles

4. Avoid leading items.

Items should be as objective and value-neutral as possible.

Does knowing that Speedplay is the highest rated extreme


sports equipment make you more confident in your purchase?
Yes
No

The tight-fisted conservatives in Congress care nothing about the


nation’s poor children.
Strongly
Strongly
1 2 3 4 5
Disagree
Agree

M KT 450 Item Development Principles

5. Avoid yes or no items.


Yes/No items can almost always be rewritten to capture
more information with no more respondent effort.

Do you have confidence in the quality of Speedplay brand


extreme sports equipment?
Yes

No
How much confidence do you have in the quality of Speedplay
brand extreme sports equipment?
None Very A
Great
at All Little Deal
0 1 2 3 4 5

7
M KT 450 Item Development Principles

5. Avoid yes or no items.


Yes/No items can almost always be rewritten to capture
more information with no more respondent effort.

Do you usually buy snacks when you see a movie?

Yes

No
How often do you buy snacks when you attend a movie
at a theatre?
Never Rarely Sometimes Often
Always

M KT 450 Item Development Principles

6. Avoid “checklist” items.


Checklist items are essentially a series of yes or no questions.

Which of the following sporting events do you enjoy attending?

football

baseball

basketball

hockey

8
M KT 450 Item Development Principles

6. Avoid “checklist” items.


Checklist items are essentially a series of yes or no questions.
How much do you enjoy attending the following sporting events?
Not Very A Great
at All Little Deal
football 0 1 2 3 4 5

baseball 0 1 2 3 4 5

basketball 0 1 2 3 4 5

hockey 0 1 2 3 4 5

M KT 450 Item Development Principles

6. Avoid “checklist” items.


Checklist items are essentially a series of yes or no questions.
How much do you like attending the following sporting
events?
Dislike Like
Very Much Very Much
football -2 -1 0 +1 +2

baseball -2 -1 0 +1 +2

basketball -2 -1 0 +1 +2

hockey -2 -1 0 +1 +2

9
M KT 450 Item Development Principles

7. Make certain scales and questions make sense together.


Scales should provide logical answers to questions.
How much do you like attending the following sporting
events?
Strongly Strongly
Disagree Agree
football -2 -1 0 +1 +2

baseball -2 -1 0 +1 +2

basketball -2 -1 0 +1 +2

hockey -2 -1 0 +1 +2

M KT 450 Item Development Principles

8. Make certain scales make sense given what’s being measured.


Does the concept have positive and negative extreme values?.

When buying clothes, how important is brand name to you?

Very Very
Unimportant Important
-2 -1 0 +1 +2

What does “very unimportant” mean?

10
M KT 450 Item Development Principles

8. Make certain scales make sense given what’s being measured.


Does the concept have positive and negative extreme values?.

When buying clothes, how important is brand name to you?

Not At All Slightly Very


Important Important Important
0 1 2 3 4 5

M KT 450 Item Development Principles

8. Make certain scales make sense given what’s being measured.


Does the concept have positive and negative extreme values?.

On the scale below, indicate your feelings toward


tobacco advertising.
Very Very
Unethical Ethical
-2 -1 0 +1 +2

What does “very ethical” mean?

11
M KT 450 Item Development Principles

8. Make certain scales make sense given what’s being measured.


Does the concept have positive and negative extreme values?.

On the scale below, indicate your feelings toward


tobacco advertising.

Not At All Slightly Very


Unethical Unethical Unethical
0 1 2 3 4 5

M KT 450 Direct versus Indirect Measures

Observable variables: You should ask directly.

Example: Asking directly about age

Please indicate your age by selecting the appropriate


category below. (Scale with age ranges.)
Example: Asking indirectly about age
Please indicate the percentage of your remaining hair
that’s not grey.

12
M KT 450 Direct versus Indirect Measures

Observable variables: You should ask directly.


Unobservable variables: You can ask directly or indirectly.

Example: asking directly about purchase satisfaction


I am satisfied with the clothes I purchased from
Elder Beerman. (scale for level of agreement)

Example: asking indirectly about purchase satisfaction


I make positive comments to friends about my recent
clothes purchase from Elder Beerman (level of agreement
or frequency).

M KT 450 Direct versus Indirect Measures

Observable variables: You should ask directly.


Unobservable variables: You can ask directly or indirectly.

Example: asking directly about purchase satisfaction


I am satisfied with the clothes I purchased from
Elder Beerman. (scale for level of agreement)

Example: asking indirectly about purchase satisfaction


I make positive comments to friends about my recent
clothes purchase from Elder Beerman (level of agreement
or frequency).

13
M KT 450 Direct versus Indirect Measures

Observable variables: You should ask directly.


Unobservable variables: You can ask directly or indirectly.

Example: asking directly about interest in fine arts.


Indicate your overall level of interest in fine art
(scale for interest level).

Example: asking indirectly about interest in the arts


In an average year, how often do you attend art exhibits?
(scale for frequency).

M KT 450 Measures and Types of Data

Scales produce four basic types of data:

nominal: classifies responses by attribute, name or type.

ordinal: arranges responses according to their strength or


magnitude
interval: arranges responses by magnitude and does so in
measurably spaced units or intervals but has no naturally
occurring zero value.
ratio: arranges responses by magnitude and does so in
measurably spaced units or intervals and has a natural value
for zero.

14
M KT 450 Measures and Types of Data

Three types are of interest to market researchers:

examples of scales producing nominal data

Are you male? female?

Indicate your make of automobile:


Ford Chevrolet Honda etc.

M KT 450 Measures and Types of Data

Three basic types of data:

examples of scales producing ordinal data

Indicate your satisfaction with your waiter or waitress.


Completely Satisfied
Somewhat Satisfied What’s the “distance”
Neutral covered by each
Somewhat Dissatisfied response category?
Completely Dissatisfied

15
M KT 450 Measures and Types of Data

Three basic types of data:

examples of scales producing ratio data

Indicate your income:


zero to $9999
10,000 to 19,999 What’s the “distance”
20,000 to 29,999 covered by each response
category?
30,000 to 39,999
40,000 or more

M KT 450 Sums and Averages

Measures can be created by mathematically combining other


measures.
Sums: A measure is calculated as the sum of two or more
other measures.

16
M KT 450 Sums and Averages

Example: Measure of “local sports enjoyment”


Measures can be created by mathematically combining other
measures. “In the past year, how many of each of the following
sporting events have you attended?
Sums: A measure is calculated as the sum of two or more
oth er High School sports
measures.
None 1-2 3-4 5 or more

Dayton Dragons None 1-2 3-4 5 or more

Dayton Bombers None 1-2 3-4 5 or more

UD Flyers None 1-2 3-4 5 or more

None: score 0
1-2: score 1
0+1+0+3
3-4: score 2
5+: score 3 =4

M KT 450 Sums and Averages

Measures can be created by mathematically combining other


measures.
Sums: A measure is calculated as the sum of two or more
other measures.

Avoid excessively lengthy sums.

Individual items should make sense as parts of the


concept being created by the sum.

Individual items being summed should have identical


(or highly similar) scales.

17
M KT 450 Sums and Averages

Measures can be created by mathematically combining other


measures.
Sums: A measure is calculated as the sum of two or more
other measures.

Averages: A measure is calculated as the sum of two or more


other measures.

M KT 450 Sums and Averages

Example: Measure of “attitude toward ad”


Measures can be created by mathematically combining other
“Please indicate your reaction to the advertisement by
measures.
circling the number between each word pair that best
Sums: measure isyour
reflects feelings.”as the sum of two or more
calculated
A oth er measures.bad 1 2 3 4 5 good
Averages: A measure isdifficult 1 2as3the
calculated 4 sum
5 easy
of two or more
oth er measures.
ineffective 1 2 3 4 5 effective
dull 1 2 3 4 5 exciting

4 + 3 + 4 + 2 = 13

13 ÷ 4 = 3.25

18
M KT 450 Sums and Averages

Measures can be created by mathematically combining other


measures.
Sums: A measure is calculated as the sum of two or more
other measures.

Averages: A measure is calculated as the sum of two or more


other measures.
Items being averaged should tap some part of the same
overall concept (more later).
Averages should not mix observable and
unobservable variables.
Scales can differ, but should have same number of response
options.

M KT 450 Validity and Reliability

Validity
In questionnaire development, validity refers to the degree to which
items accurately tap the intended concepts.

With established or verifiable measures,


validity can be easily established.

With new measures, validity may be


up to researcher judgment.

19
M KT 450 Validity and Reliability

Validity
Multi-scale items help improve validity, especially when
measuring constructs (e.g., attitudes).
Each scale
Example: attitude scale should
tap part of the same
basic concept.
bad 1 2 3 4 5 6 7 good
unreliable 1 2 3 4 5 6 7 reliable
poorly made 1 2 3 4 5 6 7 well made
weak 1 2 3 4 5 6 7 strong
low quality 1 2 3 4 5 6 7 high quality

M KT 450 Validity and Reliability

Validity
Multi-scale items help improve validity, especially when
measuring constructs (e.g., attitudes).

(un)reliable
poorly/well made

Respondent’s
True
Attitude
w
e
a
k
/
s
t
r
o
n
20
g
M KT 450 Validity and Reliability

Reliability

Reliability refers to the degree to which a measure yields consistent


results.

Reliability can be established using statistical techniques.

test-retest method

equivalent forms method

M KT 450 Validity and Reliability

Reliability: Equivalent Forms

Include on same questionnaire an “equivalent” item or items.

First (Primary) Form Second


(Equivalent) Form
bad 1 2 3 4 5 good I’m glad I purchased
this product.
unreliable 1 2 3 4 5 reliable
Strongly Strongly
poorly made 1 2 3 4 5 well
Disagree Agree
made
1 2 3 4 5
weak 11 22 3
low quality 3 4
4 55 strong
high quality

21
M KT 450 Comparing Validity and Reliability

low reliability high reliability high reliability


and questionable and high and poor
validity validity validity

22

You might also like