Roman C. answered 09/27/22
Masters of Education Graduate with Mathematics Expertise
The sampling distribution for the standard score in this case is student's t-distribution with n-1 degrees of freedom.
We need to cover the middle 0.95 probability, so the cdf must have F(-t*)=0.025 and F(t*) = 0.975, which we can look up in the t-table with df=34 or use software.
I used R-Studio by calling the routine qt(0.975,df=34).
t* = 2.032245
Now we also need the standard error:
SE = s / √n = 4.8 / √35 = 0.8113481
So the confidence bounds are x̄ ± t*SE.
x̄ + t*SE = 19.2 + 2.032245 * 0.8113481 = 20.84886
x̄ - t*SE = 19.2 - 2.032245 * 0.8113481 = 17.55114
The 95% confidence interval is approximately (17.55, 20.85)