"Here is a program that will help you to do a chi-squared test":
ChiTest:=proc(E,F)
local r,s,i;
r:=0;
for i from 1 to nops(E)
do
r:=r+((E[i]-F[i])^2)/E[i];
od;
s:=evalf(100*(1-stats[statevalf,cdf,chisquare[nops(E)-1]](r)));
lprint("The chi-squred value is",evalf(r),"where there are",
nops(E),"classes, and a probabilty of about ",
s,"percent that the null hypothesis is true and we see such a distribution.");
if
s > 5
then
lprint("In particular this test is not significant and we should not reject the null hypothesis based on it");
fi;
if
1<s and s<= 5
then
lprint("In particular this test is significant and we may reject the null hypothesis based on it");
fi;
if
s<=1 then
lprint("In particular this test is highly significant and we may reject the null hypothesis based on it");
fi;
end:ChiTest([58*.125,58*.125,58*.375,58*.375],[4,4,17,33]);