potential questions

Systematic gathering and interpretation of information about individuals or organizations..

..using statistical and analytical methods and techniques of the applied social sciences...

..to gain insight or support decision making.

Generating
-> insight to concept

Developing
-> concept to product

Monitoring
-> product to money

to gain customer insight to help organizations make better decisions.

The choice of tools depends on the situation.

Ad Hoc
(specific research issue)

Programmatic
(returning research)

if decision has already been made
or
if decision needs to be made in an instant

if there is uncertainty about what to do

-> uncertainty should be balanced against the costs of conducting MR

...it is a plan for a study that ensures to be relevant and cost efficient!

Different types:
Ambiguous problems: Exploratory R
Methods: Focus Groups, Personal Interviews, Observational studies

Somewhat defined; Descriptive R
Methods: One-Time-Survey, Logitudinal survey, panel studies

Clearly defined: Causal R
Methods: Laboratory studies, field studies, designs w/ high control of individual variables

Research projects
-> to adresss specific research projects

Information systems
-> continuous monitoring of consumers and market behaviour

Marketing Manager
has to gather information, sees only obvious and observable problems (surface level)

Market Researcher
turns decision problems into research problems...
looks under the surface (root of problem)

collected for specific research problem
so called "field research"

advantages:
recent
specific for research problem
proprietary

disadvantages:
more expensive
takes longer

collected for other purposes than the specific research problem
so called "desk research"

advantages:
cheaper
bigger sample size
quick to access

disadvantages:
may be outdated
may not fit research problem
may have errors

internal
..internal issue within company

...data already collected -> now used for r. problem

external
..data about e.g. competitors

...belongs to other companies (costs)

1. Free interpretation

2. Sequential analysis

3. Coding

4. Qualitative content analysis

Open Coding
(give concept codes to incidents -> fit data!)

Axial Coding
(see connection, sort them, link everything)

Selective Coding
(everything brought together)

...you know possible values your variables take on before starting research

Methods
surveys, experiments

Values:
Age (in years)
Income (in $/€)

...you don't know possible values your variables take on before starting research

Methods:
Focus Groups, Observation studies

"Values":
holiday experience (leaves open-end)

Flexible and systematic standards for collection and analysis of qualitative empirical data

AIM:
deriving inductive conclusions
checking these conclusions for validity

simultaneous involvement of multiple respondents
interactive, structured, and spontaneous small group discussions

ADVANTAGES:
Information
Interaction
Efficiency

DISADVANTAGES:
Process issues
Generalisability

"Watching what people do"

ADVANTAGES
Actual Data
Unconscious behaviour

DISADVANTAGES
No additional values

Prolonged observation of consumer behaviour, emotional responses

AIM
understand group of interacting people

ADVANTAGES
Realism
Holistic
Observe subcultures

personal and verbal surveys using mainly open-end questions
forms: structured, unstructured

ADVANTAGES
Individuality
Deep understanding
Flexibility

(Projective Interview, Laddering Technique)

Experiment is done in a small section of marketplace. The aim is predicting sales results of some proposed marketing action

I
I
V

simulated market
works?
I
V
standard test market
works?
I
V
national rollout

Population
Group of units about which to draw conclusions from

Sample
Subset of selected cases from the population

Sample Frame
List containing all or random selection of population

simple words/ short sentences
avoid ambiguity
avoid leading questions
avoid implicit assumptions
avoid generalisations
avoid double-barraged questions
avoid negotiations

Cluster sampling
(take cluster of pop. that represent pop. - take samples of cluster)

Stratified sampling
(sort for characteristics - after sorting=strata - random samples of strata)

Simple Random Samping
(completely random selection - equal chance of being selected)

Judgemental sampling
(researcher takes informed guess - equal chance of being selected)

Quota sampling
(selecting sampling)

NON METRIC

Nominal ( Which of these do you like?)

Ordinal (Rank these)



METRIC

Interval (Please indicate on scale)

Ratio (On a scale of 1-100)

is it free from error?

What determines validity?

affects in a predictable way!
e.g. measuring with a shit ruler

affects in a not systematic way
e.g. measuring with a weird ruler

Noncoverage
fail to include units

Nonresponse
Not-at-Homes, Refusals

Data collection error
refuse to answer

Office processing error
error while coding

single-item is more efficient but reduces quality of measurement
What are benefits of single-item-measurement vs. multi-item-measurement?

Laboratory Experiment

conditions result of created situation
independent variables are manipulated
other variables. can be controlled perfectly

Field experiment!

Realistic situation
Independent variables manipulated
Other variables controlled as careful as situation permits

Probability sampling

each population has a known non-zero chance of being included in the sample



Non-probability sampling

includes personal judgement of researcher. The cost of sample framing are eliminated but no way of ensuring representativeness.