Two-way ANOVA...or not?

#1
The following data is observed by a building society:

No of mortgage applications per month
Year 1989 1990 1991 1992 AVERAGE
Season
Spring 94 115 78 97 96
Summer 65 95 55 85 75
Autumn 31 44 35 46 39
Winter 50 50 32 36 42

AVERAGE 60 76 50 66 63

The building society needs to assess how reliable is the seasonal pattern from year to year, and whether variations from one year to another expain a significant part of the variability of the number of mortgage applications.

a) Do an analysis of variance test for the building society.

b) Discuss to what extent the techniques of χ2 and multiple regression could have been used to give information to the building society comparable to that in part (a).

--------------------------------------------------------------------------------------

a) I did the two-way anova, but I'm not sure that's right given the question in b). Is that right?

b) I can't find anywhere the relationship between anova and chi square or multiple regressions, and the conditions upon which I should use one or the other. Any hints?

Thanx again.