Regression Flashcards
CPM
Cost-per-thousand (CPM) = ( cost / impressions) x 1000
CTR
Clickthrough rate (CTR) = (Clicks / Impressions)
CVR
sales / clicks
CPC
Cost-per-click (CPC) = Cost / Clicks
Cost per conversion
Cost / conversion
Return on advertising spend (ROAS)
profit / cost
Conversion rate per sale
Sales / conversions
Cost per sale
Cost / Sales
alternative profit
impressions x CTR x (CVR x M - CPC)
Slope:
A unit change in X leads to β units change in Y
Intercept:
Average value of Y when X = 0
Y X
A unit change in X leads to β units change in Y
Y Log(x)
A 1% change in X leads to a β units change in Y
log(y) x
A unit change in X leads to a β % change in Y
log(y) log (x)
A 1% change in X leads to a β % change in Y
biased results when
Omitting relevant variables or incorrectly specifying the relationship between variables
How to obtain coefficients that are interpreted as elasticities?
Transforming the
variables and calculating by Natural logarithm (Ln) transformations
Natural logarithm (LN) transformations:
ln(ProductSales_i )=β_0+β_1 ln(VolumeOwned_i )
+ β_2 ln(VolumeEarned_i)
If a coefficient is an elasticity in log transformation we use an increase or a decrease in
percentage and not in percentage points
Ex.
For every 1% of increase in (…) leads to an increase of pvalue% in …
interaction term
Y_i=β_0+β_1 X_(1,i)+β_2 X_(2,i)+β_3 X_(1,i)×X_(2,i)
X_(1,i)×X_(2,i)
when interaction effect
when depends on or onfluence on