lecture 4 Flashcards
lurking variables
variables that is not included as an explanatory or response variable in the analysis but can affecet the interpretation of relationships between variables
also called confounding variables
sample selection bias
failing to ensure that the sample obtained is representative of the population intended to be analyzed
No control group
effectiveness compared to no intervention (eg promo strat)
Gold standard data driven marketing
anchor on a decisions that needs to be made
Finds data for a purpose
Start from what is unknown
Empower decision making
Field experiments
field experimentation represents the conjuction of two methodological strategies: experimentation and field work
Key features:
1) authenticity of treatments
2) Representativeness of participants
3) real world context
4) relevant outcome measures
IN most field experiments, participants are not even conscious that they are taking part in an experiment
True experiments (3 identifiable aspects)
1) Comparison of outcomes between treatment and control
2) Assignment of subjects is to groups is done through a randomization device
3) Manipulation of treatment is under control of a researcher/ analyst
Eight steps of an experiment
1) write down a testable hypothesis (generally advocate a “no change” hypothesis
2) Decide on two or more treatments that might impact the outcome variable(s) of interest
(generally include a control treatment where nothing is changed to use as a baseline)
3) Compute how many subjects to include in the experiment
4) Randomly divide subjects (people/stores) into groups (alos need to decide on the sample size for each group)
5) Expose each group to a different treatment
6) Measure the response in terms of an outcome variable(s) for subjects in each group
(outcomes must be chosen in advance)
7)Compare responses via a (correct) statistical test
8) Conclude whether to reject or “fail to reject” your hypothesis
steps to analyze experimental data
1) be explicit on the business question you are trying to answer
2) build an understanding of the data structure
3) compute some descriptive statistics
4) visualize the data
5) Run (the correct) statistical test
6) use the resluts to inform decision making
Descriptive statistics (what descriptive stats do you want to know)
how many observations in total
how many search sessions in total
How many hhotels are int he data
Across how many countries
How many search sessions in the treatment and control groups
Type 1 error:
false positive
Type 2 error:
false negative
regression estimates from experiments allow us to
test whether treatments have effects (same as ANOVA or t-test)
estimate a magnitude of the effect sizes (and standard errors)
which out t-test and ANOVA didnt
