Question

If you have monthly data over a number of years and the intercept is not suppressed, how many dummy variables will you introduce to test the following hypotheses, briefly specify the model and describe the dummy(ies)

Answer 1

Answer:

Dummy variable trap happens  in regression when two or more dummy factors  are perfectly multicollinear  or collinear such models can't be estimated as they don't satisfy assumption of classical linear regression model.Hence,if intercept is normal  and dummy factors  has m categories,only m-1 dummy variable should be presented.

(a) All the 12 months of the year exhibit seasonal patterns.

ANSWER:

At the point when all the a year of the year display  seasonal patterns and intercept term is available in the model,then dummy factors  has 12 categories.So,use 11 dummy factors to avoid dummy factors   trap.

Accept any month as reference month,introduce 1 dummy variable for other 11 months,which should take estimation  of 1 in relevant month and zero in other month.If intercept isn't present,introduce 12 dummy factors,one for every month.

(b) Only February,April,June,August,October,and December exhibit seasonal patterns.

ANSWER:

On the off chance that just February,April,June,August,October,and December exhibit seasonal patterns and intercept term is present,use 5 dummy factors.

Accept any month as reference month present 1 dummy  variable each for other 5 months,which take estimation  of 1 in significant  month and zero in other months.If intercept isn't  present,introduce 6 dummy factors,one for every  month.