Statistical modelling
Note that the effects of season and period are non-significant whether added before or after lagged (i.e. the effect of tsetse control) in the model (See accumulated analysis of variance).
If we click the 'Predict' button and then click 'Lagged' we get estimates of least squares means. We find that, the average weight of bulls increased from 231 to 248 kg after tsetse control was introduced. As there was no interaction of period and effect of tsetse control we can conclude that similar effects occurred both when targets and pour-on were used.
|
|
|
Regression analysis
Response variate: AM_WEIGHT
Fitted terms: Constant + Season + Period + Lagged
Accumulated analysis of variance
Change
|
d.f.
|
s.s.
|
m.s.
|
v.r.
|
F pr.
|
+ Season
|
1
|
231.08
|
231.08
|
2.56
|
0.127
|
+ Period
|
1
|
762.28
|
762.28
|
8.46
|
0.009
|
+ Lagged
|
1
|
1579.05
|
1579.05
|
17.52
|
<.001
|
Residual
|
18
|
1622.40
|
90.13
|
|
|
Total
|
21
|
4194.81
|
199.75
|
|
|
Fitted terms: Constant + Period + Lagged + Season
Accumulated analysis of variance
Change
|
d.f.
|
s.s.
|
m.s.
|
v.r.
|
F pr.
|
Period
|
1
|
762.28
|
762.28
|
8.46
|
0.009
|
Lagged
|
1
|
1680.01
|
1680.01
|
18.64
|
<.001
|
Season
|
1
|
130.12
|
130.12
|
1.44
|
0.245
|
Residual
|
18
|
1622.40
|
90.13
|
|
|
Total
|
21
|
4194.81
|
199.75
|
|
|
|
|
