Beruflich Dokumente
Kultur Dokumente
Research
Definisi
Sales analysis 92
Studies of business trends 91
Short range forecasting 89
Competitive product studies 87
Long range forecasting 87
Pricing studies 83
Benefits Drawbacks
•Directly relevant to the •Time consuming
business
•Often expensive
•Up-to-date data
obtained •Results may be
misleading if the sample
•Competitors do not size is too small, questions
have access to the are unclear or there is
findings interviewer bias
Secondary research
Steps
Distinguish between Decide on Information is Statistical Overall conclusions
the research type - budget collected manipulation of to be presented
needed e.g. according to the data collected rather than
- data sources
the plan (N.B. (e.g. regression) overwhelming
- exploratory - research or subjective
approaches it is often done statistical
- descriptive analysis of focus methodologies
- research by external
- causal firms) groups
instruments
- sampling plan
- contact methods
Comments
If a problem is The plan needs This phase is Significant Can take various
vaguely defined, to be decided the most costly difference in forms:
the results can upfront but and the most type of analysis - oral presentation
have little flexible enough liable to error according to
- written conclusions
bearing on the to incorporate whether market supported by analysis
key issues changes/ research is
iterations quantitative or - data tables
qualitative
MR Procedure 23
Quota sampling
People in the sample are chosen to reflect the proportions of
different groups in the target market e.g. 80% over 60s, 20%
under 60s when researching the market for mobility aids.
Stratified sampling
The appropriate market segment is first selected, for example,
married men, then the sample is randomly selected from this
segment.
Sampling methods
Cluster sampling
This uses random sampling from a specific area or cluster e.g.
tourist towns when researching leisure hotels.
Convenience sampling
This simply means using a sample of willing volunteers. This
method often results in bias and may only produce small samples.
Systematic sampling
Every nth person is selected from a list of the population. This can
be costly if the sample is widespread.
Reliability of sampling
The larger the sample size, the more reliable it is likely to be.
However, larger samples incur greater costs.
A 95% confidence level is usually expected when conducting
research. This means that findings are likely to be correct 19
times out of 20.
Research Design 29