619
edits
Changes
Jump to navigation
Jump to search
Women in Entrepreneurship (Issue Brief) (view source)
Revision as of 13:05, 20 February 2017
, 13:05, 20 February 2017→Tables built up to 02/07/17
no rounds, total$invested*/
==STATA regression info==
name text file "big.do"
press play - will run in stata
commands in stata:
summarize numrounds -> gives mean, skewness, kurtosis
summarize numrounds, det -> same as above except w higher level of detail
tab numrounds -> frequency count
gen -> generate a variable
gen invested = totalinvested*1000 = new variable
gen abovemeaninv =0 if totalinvested <. (where . is equal to positive infinity, it's what stata puts in when there's a missing value)
replace abovemeaninv=1 if totalivnested > mean & totalinvested <.
reg some measure of performance on some measure of women
reg numrounds womeninpower
if 2(std. err.) < coef, you're good
The relationship between numrounds and womeninpower is highly statistically significant
- R-squared - says of all the variation in the number of rounds, women in power only explains less than one percent of that variation 0.66%
if you use .log you can automatically replace it whenever you run the script
rebuild data set to contain men
need year variable - extract year from date
multivariate regression
y should be perfromance measure
x should be the explanatory variables
if you run multiple variables with , robust it will be fine
i.year = indicator that takes value 1 if it's the year it's invested
=Spring 2016=
