Estimating values and collinearity

[Franka Jussen]

(i use stata 17)

i have two variables, pisa, my dependent variable and m0, m1, m2 and m3 are my independent variables.

pisa corresponds to a students' grade in school and m0, m1, m2 and m3 (binary variables) are equal to 1 if the number of musical instruments in the home is respectively 0, 1, 2 and 3 or more (0 otherwise).

the question we need to answer is : (b) Provide a 99% confidence interval for the difference in pisa between a student that
lives in a house with 5 musical instruments compared to one with zero.
And i can't seem to figure out how i'm supposed to estimate pisa for students in a household with 5 instruments if i only have 3+.

I though of maybe using three points m0, m1 and m2 to estimate a curve [m0(0;1999), m1(1;1365), m2(2;846)] to then estimate the number of students with 3, 4 and 5 instruments at home knowing that there are 1021 students with 3+ instruments,
but i don't know how to do that, even less on stata.

m0, m1, m2 and m3 are quite collinear, i don't know to what lenght this may be useful information.

Thank you for your help!

I have more info if necessary like coefficients through reg and vif...

[Clyde Schechter]

This question has the look and feel of a homework problem. It is the policy here not to provide help with homework. To be clear, help with thesis and dissertatioin problems is happily offered. But for homework help you should consult your instructor, a teaching assistant, or, if consistent with your school's policies, your fellow students. If this is, in fact, not a homework problem, please post back.