In: Statistics and Probability
Use an appropriate design to evaluate the effect of forest type on chick survival, while accounting for variation in elevation.
1) State all relevant hypotheses (nulls and alternatives).
2) State which test was used and why you used it.
3) State conclusions after completing your analyses.
Forest type | Elevation | chick survival |
edge | 1 | 0.21 |
non-edge | 1 | 0.34 |
edge | 2 | 0.45 |
non-edge | 2 | 0.38 |
edge | 3 | 0.35 |
non-edge | 3 | 0.57 |
edge | 4 | 0.14 |
non-edge | 4 | 0.58 |
edge | 5 | 0.21 |
non-edge | 5 | 0.28 |
edge | 6 | 0.61 |
non-edge | 6 | 0.69 |
edge | 7 | 0.53 |
non-edge | 7 | 0.47 |
edge | 8 | 0.13 |
non-edge | 8 | 0.36 |
edge | 9 | 0.24 |
non-edge | 9 | 0.32 |
edge | 10 | 0.41 |
non-edge | 10 | 0.68 |
From above plot we observe that data come from normal distribution hence two way ANOVA test with one observation per cell is most appropriate here. This design is called randomised block design (RBD). Here we consider Elevation is as a blocking factor and Forest types are treatments.
3.
Two-way ANOVA: chick survival versus Forest type, Elevation
Source
DF
SS
MS F
P
Forest type 1 0.085805 0.0858050 7.63
0.022
Elevation 9 0.333205 0.0370228
3.29 0.045
Error
9 0.101245 0.0112494
Total
19 0.520255
Since both p-values<0.05 so we reject both null hypotheses
and conclude that here blocking factor (i.e. Elevation type) and
Forest types (i.e. treatments) have significant effect on chick
survival.