• Used for competitive advantage and too avoid high costs of poor decisions
The
consequences
of
marketing
research
done
poorly
or
not
done
at
all
• Branding
mistakes
• International
marketing
failures
• Ill
advised
new
product
introductions
• Nintendo-‐
underestimated
demand
• IBM-‐
overestimated
demand
• Other
examples
→ The costs versus the benefits of conducting the research: benefits outweigh
3
Acronym
The
role
and
salient
characteristics
of
marketing
research
can
be
described
by
RESEARCH
• R
ecognition
of
information
needs
• E
ffective
decision
making
• S
ystematic
and
objective
• E
xodus/
dissemination
of
information
• A
nalysis
of
information
• R
eccomendations
for
action
• C
ollection
of
information
• H
elpful
to
managers
4
research
problem
that
provides
an
appropriate
perspective
on
the
problem
2.
Specific
components
of
the
problem:
The
specific
components
focus
on
the
key
aspects
of
the
problem
and
provide
clear
guidelines
on
how
to
proceed.
• Common
errors:
problem
definition
is
too
broad
or
is
too
narrow
• Problems
arise
from
3
main
sources:
1. Unanticipated
change:
e.g.
research
into
lost
market
change,
employee
turnover
is
high
2. Planned
change
(
4
P’s):
e.g.
research
into
which
of
two
new
prototypes
is
going
to
be
more
sucessful
3. Serendipity
in
the
form
of
new
ideas
o Research
assists
in
the
development
(e.g.
Viagra
was
initially
used
for
heart
issues)
• Symptoms
are
not
problems
6
Acronym
Factors
to
consider
when
analyzing
the
environmental
context
of
the
problem
can
be
summed
up
as
PROBLEM
• P
ast
information
and
forecasts
• R
esources
and
constraints
• O
bjectives
of
the
decision
maker
• B
uyer
behaviour
• L
egal
environment
• E
conomic
environment
• M
arketing
and
technological
skills
MKTG
2305
Chapter
3
Summary:
Research
Design
Step
3:
Tasks
involved
in
a
Research
Design
What
is
a
research
design
• Research
design:
a
framework
or
blueprint
for
conducting
the
marketing
research
project
that
specifies
the
procedures
necessary
to
obtain
the
information
needed
to
structure
and/or
solve
the
marketing
research
problem.
• A
good
research
design
ensures
that
the
information
collected
will
be
relevant
and
useful
to
management
and
that
all
of
the
necessary
information
will
be
obtained-‐
process
is
effective
and
efficient.
Exploratory
Research
• Exploratory
Research:
A
type
of
research
design
that
has
as
its
primary
objective
the
provision
is
insights
into
and
comprehension
of
the
problem
situation
confronting
the
researcher
→ Used
when
management
realises
a
problem
exists
but
does
not
yet
understand
why
(e.g.
sales
are
slipping
in
a
particular
region)
→ Exploratory
research
is
conducted
on
small
and
nonrepresentative
sample,
so
the
findings
should
be
regarded
as
tentative
and
followed
by
more
defined
exploratory
research
or
conclusive
research
• It
can
be
used
for:
1. To
formulate
a
problem
or
define
a
problem
more
precisely
2. To
identify
alternative
courses
of
action
3. To
develop
hypotheses
4. To
isolate
key
variables
and
relationships
for
further
examination
7
Conclusive
Research
• Conclusive
Research:
is
research
designed
to
assist
the
decision
maker
in
determining,
evaluating
and
selecting
the
best
course
of
action
for
a
given
situation;
used
in
strategic
marketing
decisions
• It
is
based
on
the
assumption
that
the
researcher
has
an
accurate
understanding
of
the
problem
at
hand.
• The
objective
is
to
test
specific
hypotheses
and
examine
specific
relationships
• Can
move
between
exploratory
and
conclusive
research
or
do
both
at
once
e.g.
the
findings
from
secondary
data
and
focus
groups
(exploratory
research)
were
further
tested
in
a
telephone
survey
(conclusive
research)
• Conclusive
used
when:
when
managers
need
to
make
key
strategic
decisions,
and
need
to
generalize
the
data
to
the
market
as
a
whole
• Example:
large
national
survey
of
1000
randomly
selected
respondents
on
body
moisturiser
• Conclusive
Research
can
be
broken
down
into
two
sub-‐types:
1. Descriptive
Research
2. Causal
Research
Descriptive
Research
• Descriptive
Research:
a
type
of
conclusive
research
hat
has
its
major
objective
in
the
description
of
something-‐
usually
market
characteristics
or
functions.
Descriptive
research
goals
include:
1. To
develop
a
profile
of
a
target
market
2. To
estimate
the
frequency
of
product
use
as
a
basis
for
sales
forecasts
3. To
determine
the
relationships
between
product
use
and
perception
of
product
characteristics
4. To
determine
the
degree
to
which
marketing
variables
are
associated
• Example:
Swatch
watchers
does
an
analysis
of
its
own
sales
by
outlet
type
or
undertakes
consumer
perception
and
behaviour
studies.
• Major
types
of
descriptive
studies:
1. Sales
studies:
market
potential,
market
shares,
sales
analysis
2. Consumer
perception
and
behaviour
studies:
image,
product
usages,
advertising,
pricing
3. Market
characteristic
studies:
distribution,
competitive
advantage
• Techniques
for
data
collection
of
descriptive
research:
→ Secondary
data
analyzed
in
a
quantitative
as
opposed
to
a
qualitative
manner:
including
social
media
(discussed
in
Chapters
4
and
5)
→ Surveys:
an
interview
with
a
larger
number
of
respondents
using
a
predesigned
questionnaire
(Chapter
7)
→ Panels
(Chapters
5
and
7)
→ Observational
and
other
data
(Chapter
7)
• Cross-‐sectional
design:
a
type
of
research
design
involving
the
one
time
collection
of
information
from
any
given
sample
of
population
elements
(e.g.
How
did
the
Americans
rate
the
performance
of
Barack
Obama
immediately
following
the
election?)
→ Good
at:
representative
sampling,
response
bias
→ Bad
at:
detecting
change,
large
amount
of
data
collection,
accuracy
• Longitudinal
design:
two
or
more
measurements
on
the
same
variables
are
obtained
from
a
given
group
of
respondents
at
different
points
in
time
(e.g.
How
did
the
American
people
change
their
view
of
Obama’s
performance
before
and
after
his
first
year
in
office?)
→ The
term
panel
is
used
interchangeable
with
longitudinal
design:
consists
of
a
sample
of
respondents,
generally
households
that
have
agreed
to
provide
information
at
specified
intervals
over
an
extended
period.
→ Good
at:
detecting
change,
large
amount
of
data
collection,
accuracy
→ Bad
at:
Representative
sampling,
representative
bias
8
Causal
Research
• Causal
research:
a
type
of
conclusive
research
whose
major
objective
is
to
obtain
evidence
regarding
cause-‐and-‐effect
(causal)
relationships
• Causal
design:
a
design
in
which
the
independent
variables
are
manipulated
in
a
relatively
controlled
environment
• Example:
Examining
the
effect
of
a
promotional
campaign
on
the
sales
of
Mercedes
cars-‐
simulate
a
shopping
trip,
other
factors
are
controlled
(e.g.
product
quality
and
price)
• It
is
appropriate
for
the
following
purposes:
1. To
understand
which
variables
are
the
cause
(independent
variables)
and
which
variables
are
the
effect
(dependent
variables)
of
a
phenomenon
2. To
determine
the
nature
of
the
relationship
between
causal
variables
and
the
effect
to
be
predicted
• METHOD:
experiments
Acronym
The
process
of
a
research
design
may
be
summarized
by
the
acronym
DESIGN:
• D
ata
analysis
plan
• E
xploratory,
descriptive,
causal
design
• S
caling
and
measurement
• I
nterviewing
forms:
questionnaire
design
• G
enerating
the
needed
information
• N
umber:
sample
size
and
plan
9
Secondary
Data
• Secondary
data:
data
collected
for
some
purpose
other
than
the
problem
at
hand.
• Researchers
should
favour
secondary
data
before
going
ahead
with
their
own
primary
data-‐
only
if
the
secondary
data
sources
have
ben
exhausted
or
yield
marginal
returns
• Uses
of
secondary
data
include:
1. Identify
the
problem
2. Better
understand
and
define
the
problem
3. Develop
an
approach
to
the
problem
4. Formulate
an
appropriate
research
design
(e.g.
by
identifying
the
key
variables)
5. Answer
certain
research
questions
and
test
some
hypotheses
6. Interpret
primary
data
with
more
insight
• Advantages:
useful
for
estimating
demand,
monitoring
the
environment,
developing
business
intelligence
systems,
segmenting
and
targeting.
• Disadvantages:
limited
by
the
degree
of
fit
with
the
current
research
problem,
data
accuracy,
methods
used
to
collect
data
may
not
be
compatible
with
current
situation,
compatibility
of
units
of
measurement
or
timeframe.
• A
classification
of
secondary
data
>>
Acronym
The
criteria
used
for
evaluating
secondary
data
can
be
described
by
the
acronym
SECOND
• Specifications:
methodology
used
to
collect
the
data
11
Acronym
The
salient
characteristics
of
syndicated
data
may
be
described
by
the
SYNDICATED
• Surveys
• Yield
data
of
known
commercial
value
• Number
of
clients
use
the
data
• Diary
panels
• Institutional
services
• Cost
is
low
• Audits
• Timely
and
current
• Electronic
scanner
services
• Data
is
combined
from
difference
sources:
single-‐data
source
Direct (nondisguised):
Focus
Groups
• Focus
groups:
an
interview
conducted
by
a
trained
moderator
among
a
small
group
of
respondents
in
an
unstructured
and
natural
manner
(e.g.
focus
group
discussion
guide
for
smart
phone,
or
consumer
response
to
Samsung’s
newest
phone)
• Main
purpose
is
to
gain
insight
on
issues
of
interest
to
the
researcher
by
listening
to
a
group
of
people
from
the
appropriate
target
market
• Value
lies
in
obtaining
rich
information
from
free
flowing
discussions
• Focus
groups
are
the
most
important
qualitative
research
procedure
• Synonymous
with
qualitative
research
• Conducting
a
focus
group
requires:
1. Design
the
focus-‐group
environment
2. Recruit
and
select
focus
group
participants
3. Select
a
moderator
4. Prepare
the
discussion
guide
5. Conduct
the
group
interview
6. Prepare
the
focus-‐group
report
Characteristics
of
Focus
Groups
Group
Size
6-‐8
respondents
Group
Composition
Homogenous:
same
gender,
age
group,
user
category
;
respondents
prescreened
Physical
Setting
Relaxed,
informal
atmosphere
Time
Duration
1
to
3
hours
Recording
Use
of
audio
and
video
recording
Moderator
Observational,
interpersonal
and
communicational
skills
of
moderator
are
crical
• Moderator
is
key
role:
→ Well
trained
in
interpersonal
communication
→ Kind
but
firm
→ Permissive,
sensitive
→ Encourage
involvement/understanding
→ Flexible
→ Have
a
good
understanding
of
objectives/background
• Advantages:
14
In-‐Depth
Interviews
• Depth
interviews:
an
unstructured,
direct
personal
interview
in
which
a
single
respondent
in
probed
by
a
highly
skilled
interviewer
to
uncover
underlying
motivations,
beliefs,
attitudes
and
feelings
on
a
topic
• Like
focus
groups,
depth
interviews
are
an
unstructured
and
direct
way
of
obtaining
information.
Unlike
focus
groups,
however,
depth
interviews
are
conducted
on
a
one-‐one
one
basis,
these
interviews
typically
last
from
30
minutes
to
more
than
an
hour.
• As
the
interview
progresses,
the
type
of
questions
asked,
the
proves
used
and
the
wording
of
the
question
depends
on
the
answers
received
• Example:
how
shoppers
view
the
Internet
shopping
experience
• The
success
of
depth
interviews
rests
with
the
skills
of
the
interviewer
who
should:
1. Avoid
appearing
superior
and
put
the
respondent
at
ease
2. Be
attached
and
objective
yet
personable
3. Ask
questions
in
an
informative
manner
4. Not
accept
brief
“yes”
or
“no”
answers
5. Probe
the
respondent
• Two
basic
types:
1. Nondirective
interviews
2. Semi-‐structured
or
focused
Individual
Interviews
• Nondirective
interviews:
Maximum
freedom
to
respond
(so
long
as
discussion
is
related
to
area(s)
of
research
interest)
• Semi
structured
or
focused
individual
interviews:
Covers
a
specific
list
of
topics
or
sub-‐areas
(e.g.
timing,
exact
wording,
and
time
allocated
to
each
question
usually
predetermined)
• Advantages:
→ Can
uncover
deeper
insights
about
underlying
motivates
than
focus
groups
→ Can
attribute
the
responses
directly
to
the
respondents
→ Free
exchange
of
information-‐
focus
groups
may
inflict
social
pressure
to
conform
→ Probing
allows
the
information
on
complex
issues
to
be
found
• Disadvantages:
→ Need
a
highly
trained
interviewers-‐
expensive
and
hard
to
find
→ Results
are
susceptible
to
the
interviewers
influence,
quality
of
results
is
dependent
on
their
skill
→ Hard
to
decipher
the
results
→ Time
Consuming
→ Small
sample
may
not
be
representative
of
the
larger
population
Characteristic
Focus
Group
In-‐depth
Interview
Group
synergy
and
dynamics
+
-‐
Peer
pressure/
group
influence
-‐
+
Client
involvement
+
-‐
Generation
of
innovative
ideas
+
-‐
In-‐depth
probing
of
individuals
-‐
+
Uncovering
of
hidden
motives
-‐
+
Discussion
of
sensitive
topics
-‐
+
15
Association
• Association
techniques:
a
type
of
projective
technique
in
which
the
respondent
is
presented
with
a
stimulus
and
asked
to
respond
with
the
first
thing
that
comes
to
mind.
• Word
association:
a
projective
technique
in
which
respondents
are
presented
with
a
list
of
words,
one
at
a
time.
After
each
word
is
presented,
respondents
are
asked
to
give
the
first
word
that
comes
to
mind.
→ Words
of
interest
are
called
test
words
→ List
includes
neutral,
filler
words
to
disguise
purpose
of
study
• Analysed
by
calculating:
1. Frequency
of
using
same
word
2. Latency
• Response
latency:
the
longer
a
individual
hesitates
before
answering,
the
higher
the
assumed
level
of
involvement
with
the
subject
• A
non-‐response
is
thought
to
indicate
the
highest
level
of
emotional
involvement,
because
these
people
are
too
involved
to
be
able
to
respond
in
a
short
time.
• E.g.
Thoughts
that
come
to
mind
when
one
says
IKEA
are
jotted
down
by
the
researcher
Completion
Techniques
• Completion
techniques
Requires
the
respondent
to
complete
an
incomplete
stimulus
situation
• Sentence
completion:
a
projective
technique
that
requires
the
respondent
to
complete
a
number
of
incomplete
sentences
using
the
first
word
or
phrase
the
comes
to
mind.
• More
directed
than
associations,
but
not
as
disguised
• E.g.
Ikea
reminds
me
of….
People
who
shop
at
Ikea
are…
My
friends
think
Ikea
is….
Friend
visiting
from
overseas
thinks
Ikea
is….
• Story
competition:
respondents
are
provided
with
part
of
a
story
and
are
required
to
give
their
conclusion
in
their
own
words.
16
Construction
Techniques
• Construction
techniques:
requires
the
respondent
to
construct
a
response-‐
in
the
form
of
a
story,
dialogue
or
description
• Picture-‐
response
technique:
the
respondent
is
shown
a
picture
and
asked
to
tell
a
story
describing
it;
an
analogy
e.g.
the
cola
world.
Respondents
make
up
a
story
→ These
stories
are
analyzed
in
an
attempt
to
identify
themes
reflecting
the
individuals
perceptual
interpretation
of
pictures
• Cartoon
tests:
cartoon
characters
are
shown
in
a
specific
situation
related
to
the
problem.
The
respondents
are
asked
to
indicate
the
dialogue
that
one
cartoon
character
might
make
in
response
to
the
comments
of
other,
usually
unspoken
thoughts
→ 1
Is
given
response
is
given-‐
participant
is
asked
to
fill
in
the
other
(e.g.
lets
get
some
clothes
from
target,
other
person
fills
in
response)
→ Tends
to
maximize
the
candid
nature
of
responses
• Used
to
evaluate
perceptions
and
attitudes
towards
topic
Expressive
Techniques
• Expressive
techniques:
the
respondents
are
presented
with
a
verbal
or
visual
situation
and
asked
to
relate
the
feelings
and
attitudes
of
other
people
to
the
situation
• Role
playing:
respondents
are
asked
to
play
the
role
or
assume
the
behaviour
of
someone
else
• Third
Person
Technique:
respondent
is
presented
with
a
verbal
or
visual
situation
and
asked
to
relate
the
beliefs
and
attitudes
of
a
third
person
to
the
situation
(such
as
their
parent/neighbor)
• May
be
used
when
questions
are
embarrassing/
answers
aren’t
socially
desirable
• Assumes
individual
will
reveal
personal
beliefs/attitudes
while
describing
other/situation
• Reduces
social
pressure
errors
Applications:
• Projective
techniques
are
used
less
than
focus
groups
or
in-‐depth
interviews
• Word
association
is
the
most
commonly
used
for
testing
brand
names,
measure
attitudes
about
particular
products,
brands,
packaging
or
advertisement
• Their
usefulness
is
enhanced
when:
1. Sensitivity
of
the
subject
matter
is
such
that
respondents
might
not
be
willing
or
able
to
answer
honestly
to
direct
questions
2. To
uncover
subconscious
motives,
beliefs,
or
values,
providing
deeper
insights
and
understanding
as
part
of
exploratory
research
3. Administered
and
interpreted
by
trained
interviewers
who
understand
their
advantages
and
limitations.
17
Acronym
Key
characteristics
of
a
focus
group
can
be
described
by
FOCUS
GROUPS:
• Focused
on
a
particular
topic
• Outline
prepared
for
discussion
• Characteristics
of
the
moderator
• Unstructured
• Size:
8
to
12
participants
• Group
composition:
homogenous
• Recorded:
audio
and
video
recording
• Observation:
one
way
mirror
• Undisguised
• Physical
setting:
relaxed
• Several
sessions
needed:
1-‐
3
hours
each
18
In
Home
Complex
questions
can
be
asked
Low
control
of
field
force
Personal
Good
for
physical
stimuli
High
social
desirability
Very
good
sample
control
Potential
for
interviewer
bias
High
quantity
of
data
Most
expensive
Very
good
response
rate
Some
samples
(e.g.
high-‐crime
areas)
Longer
interviews
can
be
done
may
be
difficult
to
access
May
take
long
to
collect
the
data
Mall
intercept
Complex
questions
can
be
asked
High
potential
for
social
desirability
Very
good
for
physical
stimuli
Potential
for
interviewer
bias
Very
good
control
of
environment
Moderate
quantity
of
data
Very
good
response
rate
High
cost
CAPI:
respondent
sits
in
front
of
Complex
questions
can
be
asked
High
social
desirability
a
computer
terminal
and
Very
good
for
physical
stimuli
Moderate
quantity
of
data
answers
a
questionnaire
on
the
Very
good
control
of
environment
High
cost
screen
usually
in
the
form
of
Very
good
response
rate
user-‐friendly
electronic
Low
potential
for
interviewer
bias
packages-‐
used
in
shopping
malls,
trade
shows
or
conventions
Mail/Fax
interview
No
field-‐force
problems
Limited
to
simple
questions
Mail
No
interviewer
bias
Low
sample
control
for
cold
mail
Moderate/high
quantity
of
data
No
control
of
environment
Low
social
desirability
Low
response
rate
for
cold
mail
Low
cost
Low
speed
Mail
Panel:
a
large
and
nationally
No
field-‐force
problems
Limited
to
simple
questions
representative
sample
of
No
interviewer
bias
No
control
of
environment
households
that
have
agreed
to
High
quantity
of
data
Low/Moderate
speed
periodically
participate
in
mail
Low
social
desirability
questionnaires,
product
tests
and
Low/moderate
cost
survey
research.
Good
sample
control
Electronic
Email
No
interviewer
bias
Low
sample
control
Low
cost
No
control
of
environment
Low
social
desirability
Low
response
rate
High
speed
Moderate
quantity
of
data
Contract
hard-‐to-‐reach
Security
concerns
respondents
Internet
No
interviewer
bias
Low
sample
control
Low
cost
No
control
of
environment
Low
social
desirability
Low
response
rate
Very
high
speed
Moderate
quantity
of
data
Visual
appeal
and
interactivity
Personalised,
flexible
questioning
Contact
hard-‐reach
respondents
Electronic
surveys
software
and
Can
perform
tasks
such
as:
sites
• Personalization
• Incorporate
complex
skip
patterns
• Randomize
response
choices
• Consistency
checks
• Add
new
response
categories
as
the
interviewing
progresses
Free/low
cost
survey
sites
Allow
users
to
create
and
file
their
own
surveys
for
free
or
at
low
cost.
Users
provide
respondents
at
their
own
discretion.
20
Observation
• Observation:
The
recording
of
behavioural
patterns
of
people,
objects
and
events
in
a
systematic
manner
to
obtain
information
about
the
phenomenon
of
interest
• Physical
actions,
body
language,
tracking
eye
movement,
temporal
actions,
amount
of
wear
on
floor
in
front
of
exhibits,
license
plate
surveys,
tracking
time
people
fixate
on
an
object
Personal
observation:
• Personal
observation:
a
trained
observer
collects
the
data
by
recording
behaviour
exactly
as
it
occurs
(e.g.
observing
traffic
counts
in
a
department
store)
1. Humanistic
inquiry:
form
of
personal
observation
in
which
the
researcher
is
immersed
in
the
system
under
study
(e.g.
asks
another
shopper
on
advice
for
which
brand
to
choose)
• Advantage:
→ Highly
flexible
method
because
the
observer
can
record
a
wide
variety
of
phenomena
→ Highly
suitable
in
for
use
in
natural
settings
(e.g.
a
regular
sales
meeting)
• Disadvantage:
→ Information
is
recorded
after
observation-‐
leading
to
high
observation
bias
→ Doesn’t
tell
you
why
people
are
behaving
like
that
Mechanical
observation
• Mechanical
observation:
an
observational
research
strategy
in
which
mechanical
devices
rather
than
human
observers,
record
the
phenomena
being
observed.
• Devices
may
not
require
respondents
direct
participation
• Example:
Peoplemeter
is
attached
to
a
television
and
continually
records
not
only
the
channels
the
television
is
tuned
to
but
who
are
watching.
• Advantages:
→ Low
observation
bias
→ Low-‐to-‐medium
analysis
bias
• Disadvantages:
→ Can
be
intrusive
→ Not
always
suitable
in
natural
settings
→ Good
for
identifying
unconscious
behaviour
patterns
or
behaviours
that
individuals
might
be
unwilling
to
discuss
honestly.
→ Best
technique
for
phenomena
that
occur
frequently
or
are
of
short
duration-‐
in
these
situations
it
may
cost
less
and
be
faster
than
survey
methods.
• Observational
disadvantages:
→ Provides
insight
into
what
behaviour
is
occurring
but
not
why-‐
attitudes,
motivations
and
values
are
all
lost
→ Highly
personal
behaviours
are
not
available
for
interaction
(e.g.
personal
hygiene)
→ Observer
might
overlook
important
features
through
selective
observation-‐
threatens
integrity
→ Can
be
adopted
only
for
frequent
behaviours
of
short
duration
as
behaviours
occurring
infrequently
or
long
term
are
too
expensive
and
time
consuming
to
record
• Best
to
view
observation
as
a
complement
to
survey
methods
rather
than
in
competition
Acronym
The
classification
of
survey
methods
by
administration
may
be
described
by
METHODS:
• Mail
panels
• Electronic
interviews
• Telephone
interviews
• Home
(in
home
personal)
interviewing
• On-‐site
mall
interviews
• Direct
mail
interviews
• Software
for
CATI-‐CAPI
• The
interviewing
method
can
affect
the
design
of
the
questionnaire
• Again
consider
the
respondent
group
• Consider
issues
such
as:
→ Personal:
Face-‐to-‐face
give
opportunity
for
feedback,
clarification,
can
be
lengthy,
and
incorporate
visual
aids-‐
as
respondent
cant
see
questionnaire
questions
must
be
short
and
simple
→ Interview
administered
questionnaire:
written
in
conversational
manner
→ Mail
and
Electronic
questionnaires
are
self
administered
thus
questions
and
instructions
must
be
simple
and
thorough
→ In
CATI,
CAPI
or
internet
and
social
media
surveys
the
computer
guides
the
respondent
through
complex
skip
patterns
and
can
incorporate
randomization
of
questions
to
eliminate
order
bias
Step
4:
Design
the
question
to
overcome
respondents
inability
to
answer
Is
the
respondent
informed?
→ Answers
are
only
meaningful
if
the
respondent
is
informed
(e.g.
Husband
may
be
unequipped
to
answer
questions
on
grocery
shopping
expenditure)
→ Willing
and
able
to
answer
→ Use
filter
questions
when
the
topic
requires
specialized
experience
or
knowledge:
initial
questions
in
a
questionnaire
that
screen
potential
respondents
to
ensure
they
meet
the
requirement
of
the
sample.
→ Offer
‘don’t
know’
response
options
• Recall
the
information?
(e.g.
grocery
spend
wk/yr)
→ Omission/Telescoping/Creation
→ Recall
is
affected
by
the
event,
the
time
elapsed,
cues
that
aid
memory
→ Questions
can
be
designed
to
aid
recall
and
unaided
recall
→ Aid
recall:
lists
a
number
of
soft
drink
brands
then
asks
which
ones
were
advertised
on
tv
last
night
→ Unaided
recall:
Which
brands
of
soft
drink
do
you
remember
being
advertised
on
tv
last
night.
• Able
to
articulate
response?
→ Difficult
to
describe
the
ideal
atmosphere
of
a
store
for
example
→ Provide
visual
aids,
maps,
verbal
descriptions,
diagrams
• Will
the
respondents
give
the
information?
Depends
on:
→ Effort
required-‐
approximation
(e.g.
what
was
bought
during
the
shopping
visit)
→ Legitimate
response:
o Relevancy
of
question
o Appropriate
context?
→ Sensitive
information
(embarrassing
or
threating)
o Includes
money,
political/religious
beliefs,
involvement
in
accidents/crimes
Step
5:
Design
the
Questionnaire
to
overcome
the
respondents
unwillingness
to
Answer
• Reduce
the
effort
required
by
respondents
(tick
instead
of
list)
• Explain
to
participants
why
questions
that
don’t
seem
to
have
a
legitimate
purpose
are
in
the
questionnaire
and
explain
the
context
of
the
survey-‐
people
are
more
likely
to
object
to
questions
that
don’t
seem
to
serve
a
legitimate
purpose
(e.g.
demographics
in
a
cereal
packaging
questionnaire)
• Handling
sensitive
topics:
Put
these
questions
at
the
end
and
ease
into
them
and
use
categories
instead
of
asking
for
specific
figures
(e.g.
income)
• Increasing
response
rates
→ Incentives,
good
cover
letter,
make
the
survey
more
personalized
(e.g.
attach
tea
bag
for
a
tea)
→ Follow
up
on
their
survey
23
→ Which
brands
or
brands
of
toothpaste
have
you
personally
used
at
home
during
the
past
month?
In
the
case
of
more
than
one
brand,
please
list
all
the
brands
that
apply?:
RIGHT
• Use
simple
words:
avoid
technical
jargon
• Use
unambiguous
words
→ Avoid
ambiguous
words
and
questions
and
instead
state
specific
frames:
in
question
or
answer
e.g.
How
often
do
you
listen
to
country
music
in
a
month?
→ Use
specific
time
frames
such
as
3-‐4
times
rather
than
occasionally,
sometimes,
regularly
etc.
• Avoid
leading
questions:
giving
respondents
a
clue
about
how
they
should
answer
→ Acquiescence
bias:
bias
resulting
from
some
respondents
tendency
to
agree
with
the
direction
of
the
leading
question
(yea-‐saying)
→ E.g.
Is
Colgate
your
favourite
toothpaste
(WRONG)
vs.
What
is
your
favourite
brand
of
toothpaste
(RIGHT)
• Avoid
implied
alternatives:
→ One
that
is
not
expressed
in
the
options
→ Depending
on
how
the
question
is
worded,
responses
differ
• Avoid
implied
assumptions:
implied
assumptions
given
about
the
consequence
• Avoid
double
barreled
questions:
where
two
responses
are
called
for,
unsure
which
one
is
right
• Avoid
generalizations
and
estimates
25
A
Question
should:
1. Define
the
issue:
who,
what,
when,
where,
why
and
way
2. Use
Ordinary
words:
words
should
match
the
vocabulary
level
of
respondents
3. Use
unambiguous
words:
avoid
usually,
normally,
regularly,
frequently
and
sometimes
4. Avoid
leading
or
biasing
the
responding:
do
not
bias
or
clue
the
respondent
to
what
the
answers
should
be
5. Use
dual
statements:
statements
should
be
positive
and
negative
Acronyms
The
objectives
and
steps
involved
in
developing
a
questionnaire
may
be
defined
by
the
acronym
QUESTIONNAIRE:
Objectives:
• Q
uestions
that
the
respondents
can
answer
• U
plift
the
respondent
• E
rror
elimination
Steps:
• S
pecify
the
information
needed
• T
ype
of
interviewing
method
• I
ndividual
question
content
• O
vercoming
inability
and
unwillingness
to
answer
• N
onstructured
versus
structured
questions
• N
on
biased
question
wording
• A
rrange
the
questions
in
proper
order
• I
dentify
form
and
layout
• R
eprodction
of
the
questionnaire
• E
liminate
bugs
by
pretesting
26
The
guidelines
for
question
wording
may
be
summarized
by
the
acronym
WORDS
• W
ho,
what,
when,
where,
why
and
way
• O
bjective
questions:
avoid
leading
questions
• R
egularily,
normally,
usually
etc.
should
be
avoided
• D
ual
statements
(positive
and
negative)
• S
imple,
ordinary
words
The
guidelines
for
deciding
on
the
order
of
questions
may
be
summarized
by
the
acronym
ORDER
• O
pening
questions:
simple
• R
udimentary
or
basic
information
should
be
obtained
first
• D
ifficult
qustions
toward
the
end
• E
xamine
the
influence
on
subsequent
questions
• R
eview
the
sequence
to
ensure
a
logical
order
27
28
Scale
Characteristics
• Description:
unique
labels
or
descriptors
used
to
designate
each
value
of
the
scale
(e.g.
1=
female)
• Order:
The
relative
sizes
or
positions
of
the
descriptors;
its
is
denoted
by
descriptors
such
as
greater
than,
less
than
or
equal
to
(e.g.
preference
for
Nike
is
greater
than
preference
for
Asics),
used
in
comparative
scaling
• Distance:
the
absolute
differences
between
the
scale
descriptors
are
known
and
can
be
expressed
in
units
(e.g.
5
strongly
agree
expresses
1
more
unit
of
agreement
than
4.
Agree)
• Origin:
indicates
that
the
scale
has
a
true
0
point
(e.g.
income,
weight,
sales)
Nominal
• Nominal
scale:
a
scale
whose
numbers
serve
only
as
labels
or
tags
for
identifying
and
classifying
objects.
When
used
for
29
identification
there
is
a
strict
one-‐to-‐one
correspondence
between
the
numbers
and
the
objects
• Examples:
Bank
account
number;
Rugby
jersey
numbers
• Type
of
analysis:
counting,
frequency
(mode,
%)
• Widely
used
in
research:
to
identify
respondents
(gender,
occupation),
brands,
attributed,
stores,
characteristics
etc.
• Example:
In
the
past
year
who
made
a
contribution
to
your
household
income:
A.
male
head
b.
female
head
c.
both
• Mode:
mode
is
the
most
common
value
in
the
set
of
responses
to
a
question
→ Mode
is
most
common
which
is
1
the
category
assigned
to
females
→ Valid
percent
removes
the
missing
scores
Ordinal
• Ordinal
scale:
a
ranking
scale
in
which
numbers
are
assigned
to
objects
to
indicate
the
relative
extent
to
which
some
characteristic
is
possessed.
• Thus
it
is
possible
to
determine
whether
an
object
has
more
or
less
of
a
characteristic
than
some
other
object
• Indicates
order
and
identity
• Can
determine
the
relative
position
(greater
than
or
less
than),
but
not
‘how’
different
things
are
→ Ranking
of
teams
in
a
tournament,
runners
in
a
race
(cant
measure
distance
between
st
values:
1
place
may
have
been
4
seconds
nd rd
ahead
of
2
who
was
5
seconds
ahead
of
3 )
• Types
of
analysis:
mode
and
median
(percentiles)
• Use
in
research:
to
measure
attitudes,
opinions,
perceptions,
preferences
• Example:
in
the
past
year
what
was
your
annual
household
income
before
taxes?
1.
Less
than
$20,001,
2.
$20,001
to
$50,000…..
5.
More
than
$150,000
• Median:
median
occurs
where
half
of
the
data
is
above
the
statistic
value
and
half
is
below
rd
→ Median
(mid
point)
falls
in
3
which
represents
the
3
category
(35,000-‐49,9999)
50%
culmative
percent
in
→ Use
median
for
incomes,
salaries
etc.
as
the
median
isn’t
effected
by
outliers
whilst
mean
isn’t
Interval
• Interval
Scale:
a
scale
in
which
the
numbers
are
used
to
rank
objects
such
that
numerically
equal
distances
on
the
scale
represent
equal
distances
in
the
characteristic
being
measured.
• A
scale
in
which
the
intervals
between
the
numbers
are
meaningful
(example=
temperature)
→ Numbers
are
used
to
evaluate
objects
→ Appear
at
numerically
equal
distances
on
the
scale
→ They
represent
equal
distances
in
the
characteristic
being
measured
→ Enable
differences
to
be
compared
→ Example:
rating
scales
→ Zero
point
is
arbitrary
(unconstrained)
• Types
of
Analysis:
mode,
median,
mean
(range,
std
dev.)
• Use
in
Research:
for
measuring
attitudes
• Example:
in
the
past
year
what
was
your
annual
household
income
before
taxes?
→ Circle
from
1-‐7
1
being
much
below
average.
4
being
average
and
7
being
much
above
average
• Mean:
mean
is
the
arithmetic
average
of
all
the
data
responses
30
Ratio
• Ratio
scale:
this
is
the
highest
level
of
measurement.
• Allows
the
researcher
to:
→ Identify/classify
objects
AND
→ Compare
the
objects
(rank/order
them)
AND
→ Compare
intervals/differences
• Contains
all
the
properties
of
nominal,
ordinal
and
interval,
AND
an
absolute
zero
point
e.g.
height,
age,
money
• It
is
also
meaningful
to
compute
ratios
of
scale
values.
• Zero
point
is
fixed:
20
kgs
is
twice
as
heavy
as
10kgs
(ratio
property)
• Type
of
Analysis:
all
types
• Use
in
Research:
sales,
costs,
market
share,
number
of
customers
• Example:
In
the
past
year,
what
was
your
annual
household
income
before
taxes?
$_________
→ Please
circle
the
number
of
children
under
18
years
of
ages
currently
living
in
your
household:
0..1……7
(If
more
than
7
please
specify(
→ In
the
past
seven
days,
how
many
times
did
you
go
shopping
at
a
retail
shopping
mall?
___
of
times
→ In
whole
years,
what
is
your
current
age?
____
years
old
Primary
Scale
Basic
Description
Scale
Common
Marketing
Permissible
characteristics
examples
examples
Statistics
Nominal
Numbers
identify
Description
Social
security
Brand
numbers,
Percentages,
and
classify
numbers,
numbers
store
types,
Mode
objects
of
football
players
gender
classification
Ordinal
Numbers
indicate
Description,
order
Quality
rankings,
Preference
Percentile,
mode
the
relative
ranking
of
teams
in
rankings,
market
median
positions
of
the
a
tournament
position,
social
objects
but
not
the
class
magnitude
of
[RANKING
scales]
differences
between
them
Interval
Differences
Description,
order,
Temperature
Attitudes,
Range,
mean,
between
objects
distance
(Fahrenheit,
opinions,
index
standard
deviation
can
be
compared,
Celsius)
numbers
zero
point
is
[RATING
scales]
arbitrary
Ratio
Zero
point
is
fixed;
Description,
order,
Length,
Weight
Age,
income,
costs,
Geometric
mean,
ratios
of
scale
distance,
origin
sales,
market
all
statistics
values
can
be
shares:
write
the
computed
actual
value
down
31
Types
of
Scaling
Comparative:
‘non-‐metric
scaling’
• Scaling
techniques
in
which
there
is
a
direct
comparison
of
stimulus
objects
with
one
another
• e.g.
comparing
two
brands
on
product
quality
• Only
ordinal
or
ranked
measurements
Non-‐comparative:
‘metric
scaling’
• Scaling
technique
in
which
each
stimulus
object
is
scaled
independently
of
the
others
• e.g.
rating
single
brands
on
product
quality
(one
at
a
time)
rather
than
comparison
to
other
object
• Most
widely
used
scaling
technique
in
market
research
• Date
are
assumed
to
be
interval
• Scales
can
be
continuous
or
itemized
Comparative Scaling
Paired
comparison
• Where
respondents
are
given
two
objects
at
a
time
and
asked
to
select
one
on
some
criterion
• Data
obtained
is
ordinal
in
nature
• Most
widely
used
comparative
scaling
technique
• Used
the
stimulus
objects
are
physical
products
• Presents
to
company
the
frequency
of
first
preferences
• Number
of
comparisons=
[n
(n-‐1)/2]
(e.g.
4
brands=
6
evaluations,
5
brands=
10
evaluations)
• Disadvantages:
→ May
force
people
to
choose
→ People
can
be
indifferent
to
categories
→ Only
produces
ordinal
level
data
→ Can
produce
an
unwieldy
amount
of
comparisons
(10
jeans=
45
comparisons)
→ The
order
in
which
the
alternatives
are
presented
might
bias
the
results
→ Consumers
may
prefer
one
to
another
however
this
doesn’t
not
imply
that
they
hate
the
other
alternative
Rank
Order
• Respondents
are
given
several
objects
at
the
same
time
and
asked
to
order
them
according
to
a
criterion
• Second
most
popular
comparative
scaling
technique
• Results
in
ordinal
data
• Represents
the
shopping
environment
more
than
paired
comparisons,
takes
less
time
• Easily
understood
and
results
are
easy
to
communicate
• Commonly
used
to
measure
preferences
as
well
as
attributed
• Disadvantages:
32
→ It
is
possible
that
even
the
brand
ranked
1
is
not
liked
in
an
abosulte
sense
→ No
distance
properties
(can’t
tell
how
far
or
close
the
dislike
between
flake
and
mars
is
etc.)
Acronyms
The
four
primary
types
of
scales
can
be
described
as
FOUR
• F
igurative:
norminal
scale
• O
rdinal
scale
• U
nconstrained
zero
point:
interval
scale
• R
atio
scale
The
different
comparative
and
non
comparative
scales
can
be
represented
by
SCALES
• S
emantic
differential
scale
• C
onstant
sum
scale
• A
rranged
in
order;
rank
order
scale
• L
ikert
scale
• E
ngaged:
paired
comparison
scale
• S
tapel
scale
33
34
Semantial
7-‐point
scale
with
Brand,
product
and
Versatility
Difficult
to
construct
appropriate
bi-‐polar
Differential
bipolar
labels
company
images
adjectives
(e.g
opposite
of
calm)
Stapel
Scale
Uni-‐polar
10-‐ Measurement
of
Only
using
one
adjective
Can
be
seen
as
difficult
and
confusing
to
point
scale,
-‐5
to
attitudes
and
images
has
an
advantage
of
apply
+5
without
a
semantic
differentials
in
neutral
point
that
no
pretest
is
needed
(zero)
to
assure
the
adjectives
are
direct
opposites
Simplicity
of
construction
lends
itself
to
telephone
interviewing
The
advantages
of
this
scale
warrant
wider
applications
than
have
been
made
in
the
pasts
35
Scale
Evaluation
• A
multi-‐item
scale
should
be
evaluated
for
reliability
and
validity
(diagram
page
317)
• Total
measurement
error
is
the
sum
of
systematic
error
and
random
error
→ Systematic
error:
affects
the
measurement
in
a
constant
way
and
represents
stable
factors
that
affect
the
observed
score
in
the
same
way
each
time
the
measurement
is
made
→ Random
error:
arises
from
random
changes
that
have
a
different
effect
each
time
the
measurement
is
made.
Reliability
• Reliability
can
be
defined
as
the
extent
to
which
measures
are
free
from
random
error
→ The
extent
to
which
a
scale
produces
consistent
results
if
repeated
measurements
are
made
of
the
characteristic
• Popular
approaches
for
assessing
reliability:
→
Test-‐retest
reliability:
respondents
are
administered
identical
sets
of
scale
items
at
two
different
times
under
as
nearly
equivalent
conditions
as
possible.
The
higher
the
degree
of
similarity
the
greater
the
reliability
→ Alternative
forms
of
reliability:
an
approach
for
assessing
reliability
that
requires
two
equivalent
forms
of
the
scale
to
be
constructed,
and
then
measures
the
same
respondents
at
two
different
times
using
the
alternate
forms.
Correlation
between
alternate
forms=
reliability
→ Internal-‐consistency
reliability:
used
to
assess
the
reliability
of
the
summated
scale
and
which
refers
to
the
consistency
with
which
each
item
represents
the
construct
of
interest.
o Split-‐half
reliability:
a
form
of
internal
consistency
reliability
in
which
the
items
contributing
the
scale
are
divided
into
two
halves
and
the
resulting
half
scores
are
correlated
(e.g.
some
will
be
given
a
negative
statement
and
some
a
positive)
36
Validity
• The
validity
of
a
scale
may
be
defined
as
the
extent
to
which
differences
in
observed
scale
scores
reflect
true
differences
among
objects
on
the
characteristic
being
measured,
rather
than
systematic
or
random
errors.
• Researchers
can
assess
validity
in
different
ways:
→ Content
validity:
sometimes
called
face
validity,
involves
a
systematic
but
subjective
assessment
of
how
well
a
scale
measures
the
construct
or
variable
of
interest
→ Criterion
validity:
reflects
whether
a
scale
performs
as
expected
given
other
variables
considered
relevant
to
the
construct
(criterion
variables)
such
as
demographic
and
psychographic
variables.
→ Construct
validity:
addresses
the
question
of
what
construct
or
characteristic
the
scale
is,
in
fact
measuring-‐
requires
that
the
researcher
must
have
a
strong
understanding
of
the
theory
that
provided
the
basis
for
constructing
the
scale
o Convergent
validity:
is
the
extent
to
which
the
scale
correlated
positively
with
other
measures
of
the
same
construct
o Discriminant
validity:
the
extent
to
which
a
measure
does
not
correlate
with
other
constructs
from
which
it
is
supposed
to
differ
o Nomological
validity:
the
extent
to
which
the
scale
correlates
in
theoretically
predicted
ways
with
measures
of
different
but
related
constructs
Acronym
The
rating
scale
decisions
may
be
described
as
RATING
• R
esponse
option:
forced
versus
nonforced
• A
trractive
versus
unattracted
number
of
categories:
balanced
versus
unbalanced
• T
otal
number
of
categories
• I
mpartial
or
Neurtal
category:
odd
versus
even
number
of
categories
• N
ature
and
Degree
of
verbal
description
• G
raphics:
physical
form
and
configuration
37