are: (i) 1,1/4,1/4,1/4,1/4; and (ii) 0,3,
1,1,3. For (i), t=37.04/12.19=3.04 on 30df.
The compwise pval is:
.
002
< p < .
01 .
Using Bonferroni with 2 contrasts (it is the
number of contrasts, not the number of treat
ments that is important!!!), the compwise p
val needs to be doubled. Thus
.
004
< p < .
02
which is less than .05; thus, there is solid ev
idence against the equality of A and the av
erage of the new formulation treatments. For
(ii), t=33.55/48.76=0.69.
Using Bonferroni,
p > .
40 and there is no meaningful evidence
against the null of no linear trend for additive.
2.
(a)
ˆ
b
1
=
−
22
.
93
/
196
.
90 =
−
.
116 . This means that
for each increase in temperature of one degree
C, the rate of oxygen consumption decreases by
.116 ml/g/hr.
(b)
SSReg
=
ˆ
b
1
∗
(
−
22
.
93) = 2
.
67 . Thus, the
ANOVA table for regression can be written:
source
df
SS
MS
regr
1
2.67
2.67
error
8
.41
.051
total
9
3.08
——
The stated test is
H
0
:
b
1
= 0 vs
H
A
:
b
1
n
= 0. F=2.67/.051= 52.1
on (1,8) df. Thus,
p < .
001 and the rate de
pends strongly on the temperature.
(c) Here we focus on
This is the end of the preview.
Sign up
to
access the rest of the document.
 Fall '08
 Staff
 Null hypothesis, Bonferroni, null hypothesis states

Click to edit the document details