this looks like a chi-square goodness of fit type problem.
i'll calculate through SPSS and then calculate manually so you can see how the formulas work.
we're looking at expected versus actual.
the expected is:
30% watching each of the 3 prime time and 10% watching cable stations on a weekday night.l
actual are:
190 homes watching ABC, 100 CBS, 165 NBC and remainder watching cable.
total viewers = 550.
actual cable viewers = 95
total is 190 + 100 + 165 + 95 = 550
SPSS says that the chi-square value is equal to 58.485 with 3 degrees of freedom and a p-value of .000 which means the results are very significant and extremely likely not to be due by chance.
at the .02 significance level with 3 degrees of freedom (4 categories minus 1 = 3 degrees of freedom), the critical chi-square value is 9.8374. the probability of exceeding that chi-square value is equal to .02.
these results were achieved through SPSS.
if you were to do the calculations manually, you would have to do the following
set up a table of value as shown below:
cat actual expected (actual - expected)^2 / expected
abc 190 165 3.7879
cbs 100 165 25.606
nbc 165 165 0
cable 95 55 29.091
total 550 550 58.48485
The formula for calculating the chi-square statistic is:
chi-square = sum(o-e)^2/e
the (o-e)^2 for each of the measurements (observed and expected) is in the 4th column of the table. ths sum is at the bottom.
this value agrees with the SPSS output so it's good.
you can round the answer as required.