For tariff, consider whether it is

For tariff, consider whether it is better to use it in its linear form (i.e. to just use it as is) or the natural log of tariff (i.e. log(tariff)). • For age, consider whether it is better to use just age or also add the agesquared term. • Variables degree class and region, need to be included in the form of dummy variables appropriately. • For the regional dummies, if there are any regions for which your chosen subject has less than 20 graduates, consider grouping this region with one of its closest regions (you can do the grouping using the command ‘recode’.) • Note that it is not necessary that the tariff and age variables are included in the same form in the two regression models. For example, if you find that for Economics it is better to use tariff but for the other Subject it is better to use log(tariff), then the two models will include this variable in di↵erent form. Report the estimated regression output for the two subjects also providing the Stata commands. Also, please provide a brief explanation for your choices in terms of the form of tariff and age you decided to use (e.g. if you decided to use tariff instead of log(tariff), what was your choice based on?). If you had to do any grouping for the region variable, also provide