Determining degrees of freedom
Degrees
of freedom
[3]
for a bivariate table are computed by multiplying one
minus the number of rows, times one minus the number of columns. Specifically:
df = (#Rows – 1)(#Columns – 1)
In our example problem about the relationship between mean income and the percent black in MA census tracts, the degrees of freedom is equal to
(2 - 1) * (3 – 1) = 1 * 2 = 2.
[3] Degrees of freedom in a bivariate table can be interpreted as the number of cells in the table for which the expected frequencies are free to vary, given that the marginal totals are already set.