QUESTION 1 - ANSWER

It is possible that income and the number of household members predict the amount of money a household spends on clothing. Run a regression to see if this hypothesis finds support in our data.

The variable for amount of money spent by a household on clothing is stxcloth. To find out how size of household and household income affect this variable, run the following regression:

sort hhid
reg stxcloth totminc hhsizem if hhid~=hhid[_n-1]

(Remember that the dependent variable, or the one you're trying to explain, has to be listed first in your STATA command.)

  Source |       SS       df       MS                  Number of obs =    1028
---------+------------------------------               F(  2,  1025) =   27.28
   Model |  581182.686     2  290591.343               Prob > F      =  0.0000
Residual |  10917927.9  1025  10651.6369               R-squared     =  0.0505
---------+------------------------------               Adj R-squared =  0.0487
   Total |  11499110.5  1027   11196.797               Root MSE      =  103.21

------------------------------------------------------------------------------
stxcloth |      Coef.   Std. Err.       t     P>|t|       [95% Conf. Interval]
---------+--------------------------------------------------------------------
 totminc |   .0049883   .0006984      7.142   0.000       .0036178    .0063589
 hhsizem |    2.24158    1.08974      2.057   0.040       .1032039    4.379957
   _cons |   49.20597   6.081281      8.091   0.000       37.27278    61.13915
------------------------------------------------------------------------------

We can see from looking at the table that both explanatory variables are statistically significant, although hhsizem is almost non-significant at the p<.05. Further, the R-squared tells us that this is not a particularly powerful relationship; there are probably other variables that better predict a family's expenditures on clothing.

Back to Questions