The
solder
data frame, a design object, contains 900 observations
(rows), which are the results an
experiment varying 5
factors relevant to the wave-soldering procedure for mounting
components on printed circuit boards. The response variable,
skips
is a count of how many solder skips appeared to a visual inspection.
The
solder.balance
data frame,
contains 720 rows representing a balanced subset of
all the experimental runs. The variables (columns) are the same; it is
also a design object.
The
solder2
contains 750 observations which are the results of a similar
experiment incorporating some of the knowledge gained in the first.
The factors still have the same names and it is still a design object.
These data frames contains the following columns:
A1.5
,
A3
,
A6
,
B3
, and
B6
.
W4
,
D4
,
L4
,
D6
,
L6
,
D7
,
L7
,
L8
,
W9
, and
L9
.
1
,
2
or
3
on a board is being counted.
Comizzoli, R. B, Landwehr, J. M., and Sinclair, J. D. (1990)
Robust Materials and Processes:
Key to Reliability. AT&T Technical Journal Vol. 69, No. 6, pp. 113--128.
John M. Chambers and Trevor J. Hastie, Statistical Models in S,
Wadsworth and Brooks, Pacific Grove, CA 1992, pg. 7.
fit1 <- aov(sqrt(skips) ~ ., data = solder.balance) fit2 <- aov(sqrt(skips) ~ .^2, data = solder.balance)