page 21, equation (1.40):

the beta (β₁) should have a hat, since it is the least squares estimate that satisfies the normal equation, not the actual parameter value

page 90, equation (3.51):

the beta (β) should not have a hat

page 119, section 4.5.1:

linear independence implies that K (not K') is of full (column) rank

page 168, section 5.3, paragraph 1:

the four variables that are candidates for being dropped (based on their individual t-tests) are SALINITY, Na, K, and Zn (but not pH)

page 218, section 7.4:

all of the R-notation expressions should also have beta with a subscript of zero (β₀) after the bar (|), that is, all of those sums of squares should include adjustment for the intercept

page 298, section 9.7:

beta with a superscript of zero (β⁰) represents any (non-unique) least squares solution based on some generalized inverse; see (9.20).

page 364, section 11.2.3, expression 11.19:

s_i should be s_(i); the expression in the table of section 11.2.5 needs a similar change

page 373, section 11.3.2, expression 11.26:

s² should be "Var"

page 417, section 12.5.1, expression 12.28:

right side of equation is missing a factor of sigma-squared (σ²)

page 432, section 12.7, problem 12.26(c):

the beta with only a subscript of zero (β₀) should not have a hat

Back to Course Page