I'm doing an analysis of unemployment predictors, i.e. gender, age, occupation etc., & trying to determine the significance of the various factors. I plan to use logistic regression, however, the data I have access to gives only %age rates for each factor individually: so I would know the %age of unemployed by gender, the %age of people unemployed per occupation, per age etc., but each category would not be broken down further, so I wouldn't know the % unemployed at age x, who are also female, and who also work in a particular occupation. Is it still possible to analyse data like this and would logistic regression be the best option here?
Many thanks in advance, i'd really appreciate any help!