Aucun titre de diapositive

1
Business Research Methods Main research
approaches in marketing Prof. G. Cestre

• What type of research do we conduct ?
• What data collection methods do we use ?
• How do we analyse data ?

2
A few good references

• Marketing research, methodological foundations,
  G.A.Churchill, Jr. D. Iacobucci, Thomson, 2005
• Marketing research, D.A.Aaker, V.Kumar G.S.Day,
  McGraw-Hill, 2003
• International handbook of market research,
  R.J.Birn, ed., Kogan Page, 2001
• Handbook of interview research, context and
  method, J.F.Gubrium J.A.Holstein, ed., Sage,
  2002
• Qualitative marketing research, D.J.Carson et
  al., Sage, 2001
• Marketing scales handbook, G.C.Bruner, K.E.James
  P.J.Hensel, ed., AMA, 2001
• Multivariate data analysis, J.F.Hair et al.,
  Prentice Hall, 1998
• Using multivariate statistics, B.G.Tabachnick et
  al., Allyn Bacon, 2000

3
Major academic (peer-reviewed) journals in
marketing

• Journal of Marketing Research
• Journal of Marketing
• Journal of Consumer Research
• Marketing Science
• International Journal of Research in Marketing
• Journal of the Academy of Marketing Science
• Journal of Advertising Research
• Marketing Letters
• Journal of Business Research
• Journal of Consumer Psychology
• European Journal of Marketing
• Journal of Advertising
• Journal of Product Innovation Management

4
Marketing research process
Steps in the research process (1)
1. What research problem ? Formulate problem,
objectives 2. What data ? Identify needed
data 3. What type of research ? Exploratory,
descriptive, explanatory (causal) 4. What
sources of information ? Primary /
secondary internal / external 5. What
research approach ? Survey, observation, fo
cus groups, experimantation 6. What value of
information ? Determine cost / benefit ratio
5
Marketing research process
Steps in the research process (2)
7. What sampling method, Random, stratified,
quota, what sample size ? convenience,
representativeness, level of error 8. What
data collection tools ? Questionnaire,
mechanical observation, in depth
interview, projective techniques,
Delphi technique, .... 9. What data
analysis method ? Statistical approach
(regression, cluster, discrim, factor,
...), content analysis, .... 10. What
research results ? Validate and interprete
results
6
Examples of marketing research objectives
Issues Objectives - examples (1) Concept Early
estimate of demand, positioning Product Check
product acceptance, needed adaptations Pricing I
dentify price thressholds, test
elasticity Branding Evaluate name impact,
memorization, preference, image, mental
associations Packaging Evaluate preferences for
different options Advertising Check degree of
interest, persuasion Point of sales Establish
best location, determine image
7
Examples of marketing research objectives
Issues Objectives examples (2) Concept Show
degree of likely success of new
........ Product Check how people react to new
design of ...... Pricing Check price elasticity
of demand for ....... Branding Identify mental
associations people make with brand
X Packaging Evaluate preferences of retailers
for different options Advertising Check degree
of persuasion of humorous ad campaign P.O.S. Est
ablish best location for new store
8
Type of research and research approach
Type Exploratory Question What are
important variables ? Aim Clarify concepts,
generate ideas Approach Qualitative (group
discussions, open survey questions) Example
What do consumers look for in a sound system
? What benefits are they seeking ?
9
Type of research and research approach
Type Descriptive Question Who does what,
how ? Aim Show association between A and B,
identify trends, test hypotheses Approach
Qualitative or quantitative (survey,
observation) Example Do younger and older
consumers use differently mobile phone
functionalities (games, ring tones, messaging,
etc..) ?
10
Type of research and research approach
Type Explanatory (causal) Question What
causes what ? Why ? Aim Demonstrate impact A
has on B, test hypotheses Approach
Quantitative (experimentation) Example What
determines consumers choice of a food store ?
Why have awareness levels for most radio channels
dropped between 2000 and 2001(France) ?
11
Marketing research process
Problem statement
( Internal / external sources )
Primary Data
Secondary Data
Data collection
Quantitative Methods
Qualitative Methods
Statistical analysis When? How much? Where? Why?
How?
Content analysis Why? How? What?
Data analysis
12
Data collection
Approaches Tools / techniques Administration
Qualitative Focus groups, in-depth
interviews Special setting Delphi techniques,
completing sentences, projective techniques,
... Qual / Quant Survey structured (direct /
indirect) Face-to-face, mail, using measurement
scales or open telephone, e-mail, questions,
and unstructured (no fax, CATI /
CAPI predetermined questions) Observation
structured (direct check Human or
mechanical, facts, vs. indirect check
consequences), people-meter, pupilo- and
unstructured (look around) meter,
galvanometer,... Quantitative Experimentation,
with / without control Laboratory or
field group (one or several) some
complex (validity issues) designs
13
Qualitative approaches
Difficult to manage and to interpret
outcome Discussion groups moderator must
guide discussions without biasing
outcome Protocols individuals express
thoughts in their own words, researcher
 translates  loose expression of
thoughts Associative individuals project their
beliefs, needs, on techniques others
(projective techniques), create pictures,
associations with animals, colors, etc..
(chinese portraits)
14
Qualitative approaches An example of in-depth
interview
Teenagers-Parent Interaction in Family Decision
Making (USA)
Classification of both teenager influence and
parental response strategies, developed from
in-depth interviews with teenagers, mothers and
fathers
Sample A convenience sample of 100 families
(selected on the basis of junior high school
directories) was selected by calling families at
random. Sample of 46 boys and 54 girls between
12 and 15 years old.
Interview Process Exploratory research
Structured depth-interview with 10
open questions
Content Analysis Developing and refining
categories of analysis initially performed
by by the researcher. Subsequently, two
individuals unfamiliar with the study reviewed
category definitions and coding guidelines
coded several family transcripts their
feedback indicated that the categorizations
were complete and understandable
Source Adolescent-Parent Interaction in Family
Decision Making , Kay M. Plain, Robert E.
Wilkes, Journal of Consumer Research, September
1997
15
Qual / quant approaches
Easier to interpret data when quantitative
approach is used. However, still some
difficulties Under a survey setting - Will
question(s) asked provide needed information ? -
Can interviewee answer correctly ? - Does
interviewee want to answer correctly ? Under an
observation setting - Will observed behaviour
provide needed info ? - Does observed person
know she is being observed ? - Will the observed
person behave as usual ? Under both settings
- Cross-sectional or longitudinal research ?
(ad hoc or omnibus vs. panel data)
16
Qual / quant approaches An example of survey
questions
Do you like to go to the movies ? (Do you
actually go to the movies ? What do we want to
know ?) What was your (company) performance last
year ? (What is performance ?) How often did
you buy soft drinks during past month ? (What
are soft drinks ? How good is respondents memory
?) If you are a smoker, how much do you smoke in
a week ? (Is respondent willing to tell ? Is
he/she minimizing ?) What brands of shampoo do
you know ? (Check aided vs. unaided recall) Do
you think it is important to protect the
environment ? (Who would say no ? Socially
desirable answers) Dont you think the price of
fresh fruit is too high ? (Leading question,
respondent will tend to say yes) Also beware
of scales !!
17
Quantitative approach

• Experimentation setting Experimental designs
• Important factors
• Time frame
• What time frame should be chosen ?
• Experimental group(s)
• How many ? Who should be tested ?
• Control group(s)
• How many ? Who should be measured ?
• Manipulated variable
• What variable ? How should it be modified ?

18
Quantitative approaches An example of
experimental setting
T1 T2 Time
Experimental group O1 X O2 Control
group O3 O4
O Observation X Treatment, for example,
modifying a price
19
Data collection issues
Focus groups Survey
Observation Experimentation
x x x x
Content interpretation Sample selection Administ
ration bias Interpersonal influence Question
selection and format Non-response
rate Reliability Validity Limited
usage Choice of time and location Sample
mortality Sample sensitivity
x x x x x x
x x x x x
x x x x x x
20
Data analysis

• Under a quantitative approach, main points to be
  specified
• Does research aim at describing or explaining,
  do we want to
• show or demonstrate (make inferences) ?
• Different stats will apply
• Do we have one, two or more than two variables
  to study ?
• Different stats will apply
• What is nature of data categorical (or
  nominal), ordinal, interval,
• or metric (ratio) ?
• Different stats will apply
• N.B. Stats techniques will not be reviewed here
  !!

21
Data analysis An example of data combinations
Objective explain
What do we want to explain ? What relationship do
we want to test ?
Explanatory variable(s) Variable(s) to be
explained (independent predictors) (dependent)
Nominal Ordinal Interval / Metric
Sex, Preferred car model, Social
class Leisure activities Education
level, Preference ranking of actors Age
category or of sports activities Revenue (in
), Time spent (in hours) Age (in
years) watching TV
N.B. Different stats will apply in different cases
22
Data analysis Multivariate statistics applied
to marketing
Some of the main multivariate stats used in
marketing are
ANOVA (MANOVA), to test differences between
means Multiple regression, to check extent of
variables A, B, Cs impact on variable
D Factor analysis, to find underlying
dimensions to a whole set of variables Multi-dim
ensional scaling (MDS), to map perceived
similarities between objects (brands) Discrimina
nt analysis, to uncover which variable(s) best
discriminate between groups Cluster analysis,
to uncover variables which maximize homogeneity
within groups Conjoint analysis, to uncover the
best product profile based on evaluation of
multiple concept descriptions Structural
equations modelling (LISREL, ...), to test causal
links between variables Confirmatory factor
analysis (CFA) to test scales and validate models
23
Data analysis - Example of study (1)
The influence of family and peer groups on
consumer decisions Combiningpublic-private and
luxury-necessity dimensions with product and
brand decisions
Consumption
Private consumption
Source The influence of familial and
peer-based groups on consumer decisions ,
Childres Rao, JCR, Sept. 1992
24
Data analysis - Example of study (2)
Analyses performed

• Reliability estimate etablished for product and
  brand decisions, for the group influence scale at
  the component level and for the overall scale
• Multivariate analyses of variances (MANOVA) the
  impact of the independent variables on each
  component of peer influence was measured
• Univariate analyses of variance (ANOVA) the
  three components of peer influence were collapsed
  into one mesure.
• Test of hypotheses for peer group influence the
  mean values of peer-group influence for
  appropriate product categories were compared
• Test of hypotheses for family influence tests
  were performed on percentage of subjects
  indicating that they bought the same brand of
  product as their parents
• Independent Variables Peer vs. Family, Public
  vs. Private, Luxury vs. Necessity
• Dependent Variables Peer-group and family
  influences

25
Data analysis - Example of conjoint analysis
results
Segmentation based on benefits sought from light
fruit drinks




Seg. 3 (seeks taste)

Segments
0.085
0.491
0.185
0.242


Seg. 2 (seeks calories)



0.077
0.226
0.164
0.533
Seg. 1 (seeks sugar)


0.115
0.33
0.166
0.388
Calories
Taste
Price
Sugar
Benefits sought
Source F. Bensa M. Filser,  Les apports des
mesures conjointes à la segmentation par
avantages recherchés , Actes Congrès AFM, La
Baule, 1990