David,

I cannot follow your spread sheet for the solution here.  The spread sheets are great but sometimes it is hard to move from cell to cell to figure out what has been done what has been done.

Frank

Frank,

These are deliberately tough tests of Gujarat Chapter 5.

The chi-square variable is approximated by: (n-1)*(sample variance/hypo. population variance)
That is:
chi-squared variable = (n-1)*(sample variance/hypo. population variance)

Therefore:
hypo population variance = (n-1)(sample variance)/chi-square variable

This above implies the confidence interval:
Lower interval = (n-1)(sample variance)/(chi-square variable @ 5%)
Upper interval = (n-1)(sample variance)/(chi-square variable @ 95%)

IMO, the intuition is difficult. As this draw on ch 5, i recommend maybe “stepping back” to the initial definition of chi-square distribution before this application.

David