Unformatted text preview: Table: Pie chart: Bar graph:
• Table: Real Question: are those who saw the ad more likely to make a purchase?
• We don’t know so far.
• We only know information about each variable separately.
– A majority saw the ad
– A majority did not make a purchase
– Does this mean the ad didn’t help? Not necessarily. • We need a way to split the data into groups, and compare the results of each group.
• This requires what we call a CONDITIONAL distribution. Pie chart: Bar graph: Conditional distributions
• A conditional distribution is the marginal distribution of one variable, for one subgroup of the other variable.
• For example, you can find the conditional distribution of ad (YES, NO), given they made a purchase.
– In other words, for those that made a purchase, what % saw the ad and what % didn’t? • You can also find the conditional distribution of ad (YES, NO), given they didn’t make a purchase.
– In other words, for those that didn’t make a...
View Full Document
- Spring '10
- Purchase, Total 1, conditional distribution, conditional distributions