Flashcards in experimentation Deck (21):
generally research is...
exploratory, descriptive or causal
no preconceived ideas
describing reality but no necessarily explaining it
gives some idea of 'why'
does not mean causation! e.g. smoking linked with lung cancer: tobacco companies argued that there was another factor
strictly speaking the only was to prove a causal relationship is through...
an isolated piece fo reality, changing on variable, keeping all others the same, and observing the change n the dependent variable: this is an idealised pure experiment
what are we trying to learn?
trial rate and repeat rate
what proportion of customers will try the offer?
what proportion of those who try it will buy again?
test marketing is...
the best known form of marketing experiment... a full launch but only in a limited geographical area
we assume that the limited geographical area is...
typical of the whole market so findings can be extrapolated
test marketing can...
test a new offer without the risk/expense of a full launch
test marketing can also be used to...
experiment using the marketing mix variables as IV's to discover the effect on the DV's like sales, market share and profit
problems with test marketing
- expensive: limited area is still quite big
- low internal validity (distortion by extraneous variables like weather, local economy, sabotage by competitors)
- tips off competitors
-limited number of good test market areas (they get 'over tested' so become atypical)
- failure can damage company morale and reputation (likely to happen)
- quite an unsubtly experiment
simulated test market involves...
volunteers exposed to product advertising and promotion, then invited to shop in a simulated supermarket
advantages of a simulated test market
- good for getting trial rate
- faster and cheaper than full test market
- more lab-like conditions: greater internal validity / much tighter control of extraneous variables
- can be done entirely confidentially
disadvantages of a simulated test market
lab conditions: questionable external validity... fake supermarket with an adverts never seen before: unlikely to get realistic behaviour
mini/controlled test market
like a full test market but with limited distribution/selected channels (channels may even be specially contrived for test marketing purposes)
advantages of a mini test market
- good for getting trial and repeat rate: can actually test sales over time
- faster and cheaper then full test market
- better confidentiality
- little potential damage to company by failure
- more controlled conditions: better internal validity then full test market