A sample of 10 twins was taken In each case one twin started
Solution
here a sample of 10 twins are taken.one twin started at a community college before finishing their education at ETSU ; the other twin went directly to ETSU
let X denote the random variable denoting the points in CCTST for the twin who started community college and Y denote the same for the twin went directly at ETSU
since here the sample is of twins there must be some relation between X and Y. so here the population is
X and Y follows a bivariate normal distribution with parameters u1=E[X], u2=E[Y], s12=Var(X),s22=Var(Y), r=correlation coefficient
here null hypothesis is there is no difference in the performance. so H0: u1-u2=0
alternative is the performance is worse for students who go directly to ETSU. so u1>u2 or u1-u2>0
so H1:u1-u2>0
we have a sample of size n=10.
let Xbar be the sample mean of X and Ybar be that of Y. and let s be the sample SD of Xbar-Ybar
given that Xbar-Ybar=12 and s=5
then the test statistic is T=(xbar-ybar)*sqrt(n)/s which under H0 follows a t distribution with df n-1=9
we reject H0 iff t>talpha;9 where talpha;9 is the upper alpha point of a t distribution with df 9 and t is the observed value of T and alpha is the level fo significance
given that alpha=0.01 so t0.01;9=2.82144 t=12*sqrt(10)/5=7.589
hence t>t0.01;9
hence H0 is rejected. hence the conclusion is that at 1% level of signifiance based on the data at hand the performance is worse for students who go directly to ETSU
