Need to develop a model of variation, explaining & predicting the differences (World Values Survey has a lot of individual data). Finding out what we expect to predict subjective well-being (climate, income, marital status, employment status, etc) -- using multiple regression to see how well these variables will predict subjective well-being (getting the R2 number, which proportion of 1.00 these variables (together) lead to predicting subj_wellbeing as a dependent variable, and if high enough (.22 or above) then we go thru and find how important each is (if they were all standardized, kept constant), then they could be used to predict the overall subj_wellbeing with only that variable (if the correlation is strong enough), it should predict pretty well. That's the model. We can use this equation to predict how happy people should be from the these independent variables.
- SES variables
- unemployment
- climate
- migration
- types of industries
ESTIMATION
VERIFICATION
DATA SET
Need an individual data set (within last 10 years), a national probability sample representative of the nation, which includes the Big Five on personality and includes subj_wellbeing information, as well.
Happiness / subjective well-being
Co-morbidity study has some data
Personality data by state
lots of within state variation
Census regions
Income data from county level (3100 counties in US)
No comments:
Post a Comment