L3 Marketing Research Flashcards
(34 cards)
Data:
- implication of being quantifiable
- conseq of quantity
- does not equal important
- there’s a lot -> find the meaningful one, w a theory
Data:
- Primary VS secondary
- Internal VS external
- collected for the study at hand VS for other purposes
- collected by company itself VS by others
The 3 main steps of data-related research projects
- theoretical concept (research Q in a broader picture)
- operationalization = define n measure
- data (collection)
prevalence of unknown parameters in marketing analysis: why? solution?
Coz we lack data from other org units, or due to confidentiality. > use marketing intelligence data = census & industry reports
5=1+3+1 reasons for Marketing Research:
best one:
- Base for taking decisions
3 political PrPrCs:
- Prestige
- PR
- Consensus seeking
worst one:
- procrastinating decisions
3 Reactions to MR in short n 3 in long term,
across the 3 same dimensions
short term: gradual adaptation f comm, easy to chg features, repositioning
Long term: new biz strat, new product, new market…
How to discover consumers’ needs?
1=4+2 methods
+ 8 techniques
with qualitative marketing research e.g.:
- short term:
- Focus groups –> test new product in dev.
- In-depth interviews –> get new product ideas
- problem-centered interviews
- observation
- long term:
- Delphi method –> find (tech) long term changes
- weak signal research –> predict uncertain future
- observation
+ 6 techniques:
- Critical incidents
- Laddering (why? why? …)
- Brainstorming
- Free association
- Collages
- “Planet-game”
- text analysis
-
projective techniques:
- Thematic Apperception Tests
- word association test
- sentence completion test
- third person techniques
Research design 3 design (type)s + their methods
- Exploratory design < qualitative
- Descriptive < qualitative n quantitative (panels, surveys, secondary data)
- Experimental design < test hyp. w experiments (the only really scientific type according to some)
Lab vs field experiments: tradeoffs
Controlling all variables vs realism
weakness of descriptive research
“correlation does not imply causation”
to infer causal relationship X>Y in research, 3 conditions must be satisfied
- X & Y happen together
- X does not happen after Y
- other possible causes are excluded
Experiment - Variables, 3 types
- independent Vs > get manipulated
- dependent Vs > presumably affected n observed
- extraneous Vs > all other Vs that could presumably affect result
Experimental designs:
2 dims w 2 values each
(quasi- aka natural experiment VS experiment)
-> randomization, w control group, is NOT vs YES possible
x
(field VS laboratory)
-> realistic: high reliability = few confounding factors & construct stableness
VS
controlled: high external validity = generalizability & stability across different contexts
4=1+3 main differences bw qual & quant R
typical goal:
- goal: dev initial understanding vs recommend final decision
- properties:*
- unstructured vs structured
- non-statistical vs statistical
- Nrs: small vs big
- that’s why C-level mgrs often prefer quant R
Delphi method:
def
limit
+ 2 strengths
- 2 weaknesses
iteratively sending questionnaires to experts
but are experts really better than normal ppl in their knowledge of reality / future?
+ 2 Strengths
- No need to bring experts together physically
- No focus groups or discussion effects
- 2 Weaknesses
- Hard to retain panelists
- Future developments not always predicted correctly by iterative consensus
focus groups VS in-depth interviews:
2 vs 2
- i-d-Is are more structured
n discover more n richer info; - focus Gs yield more innovative info
n are less subject to interviewer bias
1+1+1=2 criteria to judge quantitative R
- objectivity > diff. investigators would reach same conclusions
- reliability > accuracy of measurements = w random errors, but free from systematic errors
-
construct validity >
- internal: we really measured what we wanted? is causation warranted?
- external: can results be generalised?
research 6 steps:
The basic research process includes
2+2+2: pro.des.D.S.A.R.
2 conceptual choices
- defining the problem
- determine research design
2 data collection
- design data collection method and form
- design sample and collect data
2 result elaboration
- analyze and interpret data
- prepare research report
explain 3 main research designs:
- Exploratory research design 1. generates first insight, 2. deepen understanding of a problem, and 3. clarifies issues from quantitative research – no statistical tests
- Descriptive research design describes the population as it is, includes large set of variables – w statistical tests
- Experimental research design measures causal effect of independent on dependent variables – w statistical tests
3 options for sampling (mostly in descriptive research):
- Cross-sectional design
- Single vs.
- Multiple cross-sectional design
- vs. longitudinal design
Quasi-experiment vs. (true) experiment:
Quasi-experiment vs. experiment: both can be field or laboratory, but the former lacks random assignment of treatment or control group
A/B split testing:
A/B Testing: Method used in a randomized experiment with two variants, A and B, which are the control and variation (aka treatment) in the controlled experiment
the 3 criteria to assess the quality of quantitative research:
- Objectivity ~ standardize to avoid observer bias
- reliability ~ re-test to ensure absence of error
- construct validity ~ have we really measured what we were trying to measure?
Questionnaire design process
in 9 = 2 + ( (2 + 2) + 2) + 1 steps
- Specify the information needed
- Specify the interview method
- Determine the content of individual questions
- Overcome the respondents’ inability or unwillingness to answer questions
- Choose question structure
- Choose question wording
- Arrange questions in proper order
- Identify the form and layout
- Eliminate problems by pilot-testing
