Tests whether the difference between two or more regression lines is
significant
Usage
Slope.test(...)
Arguments
...
this function takes any number of two column matrices. The
first column is the y-data (in the case of locus equations, this is the
vowel onset) and the second column is the x-data (in the case of locus
equations, vowel target).
Value
The return value consists of the following componenets:
separate
slope, intercept, r-squared, F-ratio, "d(egrees of)
f(reedom)" and "prob(ability that) line fits data" for the separate data
matrices entered.
combined
F-ratio, "d(egrees of) f(reedom)", and
"Probability of them being DIFFERENT" for the slope and for the intercept
of the combined data.
x
the combined x-data for all the
matrices.
y
the combined y-data for all the matrices.
mat
the category vectors for the combined data (consists of 1, 0
and -1).
numrows
the number of rows in each matrix.
numcats
the sum number of matrices entered.
References
see E. Pedhazur, Multiple Regression in Behavioral Research
p.436-450, 496-507.