C
orthogonal uncorrested/unique
: Helmert :
comparing
each
group mean
Assumptions into contrasts with the mean of subsequent groups
1
M K
independence
=
.
1 of observations always equal to k-1
>
-
ANOVA not Robust orthogonal t if
·
equal no
.
2
normality of residuals Trend analysis : If factor is & Scheffe >
-
complex contests
dato within each should be measured interval/rotio level A Priori
~
group on
.
M
approximately normal SSD explained by K-1 Orthogonal
7
Post-hoc
>
-
ANOUA Robust if Nk 30 thatl e st each shope. as
many
contrasts as
you want
.
3
homogeneity of va r i a n c e s positive/negative lives ellect Contests Let the dols
guide you
&
it ~
equal va r i a n c e sa c r o s s
groups positive/negative quadrastic effect E Tukey
signific
·
pos ofmeans in
ANOVA if Cone-way on
>
- Robust highest no m o re
than timesHe smallest
ANOVA
-
1 5
.
nu
homogeneous subset : of that
means
group
don't differ significantly then together
55j SS SSw Residest > tests whether there significant ANOVA result only
purpose
=
+ Ein : o re
(Yim Y) -
=
(yu - ) +
SY :n
-
Yr) significant differences between tells you that some
group
indiv indiv He means of independent differences exist not where
va r . va r .
group M va r .
.
groups ,
N
v &
↓ linear additive model :
7
How it works Y: =
M
+
2k + Eile 3 Y : - ML Interpretation
>
Mu M -
F
:Be in V
hypotheses
Ho :
M M2F . . .
: L
always
v
follow up
decision rele : If F is
large (greater Hy not No ,
st least one with post-hoc if needed
than CV or pa) , Reject Ho group mean different V
checkassumptions before interpreting results
orthogonal uncorrested/unique
: Helmert :
comparing
each
group mean
Assumptions into contrasts with the mean of subsequent groups
1
M K
independence
=
.
1 of observations always equal to k-1
>
-
ANOVA not Robust orthogonal t if
·
equal no
.
2
normality of residuals Trend analysis : If factor is & Scheffe >
-
complex contests
dato within each should be measured interval/rotio level A Priori
~
group on
.
M
approximately normal SSD explained by K-1 Orthogonal
7
Post-hoc
>
-
ANOUA Robust if Nk 30 thatl e st each shope. as
many
contrasts as
you want
.
3
homogeneity of va r i a n c e s positive/negative lives ellect Contests Let the dols
guide you
&
it ~
equal va r i a n c e sa c r o s s
groups positive/negative quadrastic effect E Tukey
signific
·
pos ofmeans in
ANOVA if Cone-way on
>
- Robust highest no m o re
than timesHe smallest
ANOVA
-
1 5
.
nu
homogeneous subset : of that
means
group
don't differ significantly then together
55j SS SSw Residest > tests whether there significant ANOVA result only
purpose
=
+ Ein : o re
(Yim Y) -
=
(yu - ) +
SY :n
-
Yr) significant differences between tells you that some
group
indiv indiv He means of independent differences exist not where
va r . va r .
group M va r .
.
groups ,
N
v &
↓ linear additive model :
7
How it works Y: =
M
+
2k + Eile 3 Y : - ML Interpretation
>
Mu M -
F
:Be in V
hypotheses
Ho :
M M2F . . .
: L
always
v
follow up
decision rele : If F is
large (greater Hy not No ,
st least one with post-hoc if needed
than CV or pa) , Reject Ho group mean different V
checkassumptions before interpreting results
