I need a Stata regression performed of Eurostat data for health issues during the Great Recession and to show the following variables:
So basically the list I provided now. I think it's in line with what we discussed, using the time between physician visits as the main variable. The as for the control variables, if you find something that you find odd/unnecessary you have full freedom to change/add.
Time between physician visits (basically sticking to your idea)
-what you said regarding adding time dimension to health, with the assumption that health deteriorates if left to itself. This variable being (in my view) a good proxy for something rather subjectively defiend such as personal health.
Healthcare expenditure (think I missed this before)
-as we wanted to gauge the time effect of neglecting physician, and its impact.
Time stayed in hospital
-measuring severity of hospitalization (especially in socialistic countries that generally discharge as soon as possible to cut medical expense). Also opens for the possibility to test whether hospitalization during increases with neglecting ones own health.
-risky behavior that can proxy for behavior people rarely admit, or have a hard time defining (depressions, some stress etc)
-similar to alcohol but tobacco consumption (by personal observation), much more often represent pure stress related tendencies
-quite self explanatory
Obesity by BMI
-not the best measure but it does apply to alot of the population, as a good measure for physical health.
-By personal experience periods with alot of work, school etc. Training progress is the best, productive people find time. By experience physical activity declines with significantly increased free time.
Gender (unsure about this one)
-Often when doing population wide studies this I feel gender is often shoehorned in due to popular demand. But here as the study also seeks to examine behavior I do feel it can add something to the model as a whole.
-I think this is the simplest way to track and monitor business cycles.
-Factor for stress and depression (especially what I mentioned earlier regarding at least Swedens social safety net). I think this is more important here than a simple unemployment statistic.
-nothing much to specify here
-Factoring in and controlling for economic progress, not just growth.
-Gauging the severity of being laid off. I.e. difference in income when employed vs unemplyed.
-By assumption people should work more during a boom, but again we want to factor out cases where more time spent working=more work related issues.
19 freelancers are bidding on average $190 for this job
do a regression analysis is: estimate the model interpret the results test the statistical assumptions of the regression correct unchecked hyptoheses