Marketing Analytics Flashcards
(45 cards)
What is a hypothese?
A prediction from a theory
What is a dependent variable?
the outcome variable, the propsed effect (example: sales)
What are the characteristics of a dependet variable?
it’s measured and not manipulated in experiments
What is an independet variable?
the proposed cause (predictor variable)
(example: type of advertising)
What are the characteristics of an independet variable?
it’s manipulated in experiments
What is the approach for qualitative research?
testing theories using language (interviews etc.)
What is the approach for quantitative research?
testing theories using numbers
What is the objective of quantitative research?
to quantify the data and generalize the results from the sample to the population of interest
What are the samples of qualitative research?
small number of non-represenative cases
What are the samples of quantitative research?
large number of representative cases
What is descriptive reserach?
aimed at capturing the structure of your data and representing it in a compact manner
What data does descriptive research collect?
user characterstics (age, gender etc.), product characteristics etc.
What is an example of analytics for descriptive research?
Cluster analysis
What is predictive research?
it uses statistical modeling techniques with the aim of predicting new or future observations based on a training data set
What data does predictive research collect?
output: making a purchase within the next month
input: age, gender, frequency of past purchases, past ad exposures etc.
What is an example of analytics for descriptive research?
Regression, decision trees, dandom forests, etc.
What is a causal inference research?
the focus is almost exclusively on research that tests causal hypotheses
What data does causal inference research collect?
outcome: making a purchase within the next month
treatment: initation of targeting campaign
confounders: for non-experimental settings (interest in product categroy etc.)
What is an example of analytics for causal inference research?
experiments with random assigment, regression, instrumental variable etc.
What is a correlation?
When we observe changes in the input variable (X) and a change in the outcome variable (Y)
What is causality?
When we change the input variable (X) and we observe the resulting changes in the outcome variable (Y)
What is a spurious correlation?
a connection between two variables that appears causal but is not
What are the conditions ofr causality?
- concomitant variation (a cause and an effext should vary together in the way predicted by the hypothesis under consideration)
- time order of occurrence (the causing event must occur befire the effect)
- absence of other possible causal factors (the factor or variable being investigated should be the only possible causal explanation)
What is systematic variation?
Differences in the dependent variable created by a specific experimental manipulation
