Marketing Research - BYU Marriott School
Download
Report
Transcript Marketing Research - BYU Marriott School
Marketing Research
Information gathered for a specific
need or want. Can be primary or
secondary
Lecture Dr. Geurts
Data Digestion
A.
Marketing research can be conducted by
an outside firm or by a department within a
company.
B.
Scientific method - hypothesis testing.
Define
problem & get prior studies that are
applicable
Design a test or experiment or get previous
results - Use data in Statistical Analysis
Regression
Write
or Analysis of Variance
& analyze results
Have results analyzed by expert
Use results for decision process
Research Process
C. Survey Research
Questionnaire - Sample (part) vs.
Census (whole)
Sources of Market Research
a.
1)
Problems
Leading question
2)
Ambiguous question - What is your
income?
3) Low response - nonrepresentative
Too
long
No inducement to reply to questionnaire
Functionally incompetent - 20% of the
population is illiterate
a.
Problems (con’t)
4)
Respondent error - people lie
5) Sampling size - the larger the better
probability
(random, cluster, stratified)
nonprobability (quota, judgment, convenience)
Frame Error
Measurement error
Sampling error!!!!!!!!! Test for significance
b.
Statistical Analysis
Tabulate
& cross tabulation
Sample
error - Occurs when statistics are
different than the population, alpha error
Regression*
- Magnitude of association,
functional relationship between two
variables. The amount of sales increase
from increasing advertising by $1000
b.
Statistical Analysis (con’t)
Correlation**
- Association strength, value
equals -1 to 0 and 0 to 1. How frequently
or consistently sales change when
advertising changes. Do changes in
advertising always increase sales. If no
change and positive change and negative
change about the same then correlation will
be 0. If sales always go up when advertising
goes up correlation will be 1. Correlation
tells us nothing about the amount of change.
Establishing Relationships
b.
Statistical Analysis (con’t)
Discriminate
analysis - Uses variables to
put people into categories. Used for
segmenting and qualifying potential buyers.
Multidimensional
scaling used to determine
product attributes
ANOVA - Analysis
of variance used in test
marketing
* Regression shows the magnitude of the relationship
** Correlation shows the strength of the relationship.
Validity
Reliability
Collect the
right type of
information
from the
respondent
Collect the
information with
high degree of
accuracy, free
of bias
Representativeness
Collect the
information from
a representative
sample of the
population
Quality Marketing Research
D. Experiments
Apply
a treatment to one group but not to a
second group
Measure
response of both groups
Statistical
test to measure if difference is
significant
Reliability
- repeat experiments give
consistent results
D. Experiments (con’t)
Validity
- measure what intend to measure
Causation
Correlation
Time
ordered- result always follows cause
No other explanation
Spurious
correlation - no relevant cause/effect
Research Attributes
E. Test Marketing - Get a few cities
which represent U.S. population
Method
Introduce
new product and vary combinations
of marketing mix variables among cities
Measure sales impact
If sales impact is significant, select the best
combination of marketing mix variables and
introduce product into a larger market
E. Test Marketing (con’t)
Advantages
Allows
you to find the best mix of marketing
variables
Able to see if product will be successful without the
cost of introducing it nationally
Provides information on consumer reaction
Disadvantages
Tip
off competition
Delay profits
F.
Panels
Get
a group of individuals that represent the
population of interest so that the panel
represents the population in characteristics
that are viewed as important.
Method
Introduce
products & evaluate
Evaluate advertising
Survey about opinions
Decision Support
Good lawyers do not allow expert
witnesses to do marketing research
for testifying in law court cases
Market Research Ad
Market Research Ad