I am wanting to do an analysis of variance on the following data (a snippet
of a larger dataset) to determine whether the factor BirdCall has a
significant effect on Responses.
Location Time BirdCall Responses
field1 morning human 3
field1 evening human 5
field1 morning recording 4
field1 evening recording 8
field2 morning human 1
field2 evening human 3
field2 morning recording 2
field2 evening recording 2
field3 morning human 4
field3 evening human 5
field3 morning recording 6
field3 evening recording 9
Since Location is fully crossed with both factors, I am wondering if I need
to specify the error term for BirdCall as follows:
aov(Responses ~ BirdCall + Error(Location/BirdCall), data=bsurvey)
Or can I simply do
aov(Responses ~ Birdcall, data=bsurvey)
Not only is there a large difference in the anova output, but the latter
method prevents me from doing the post-hoc Tukey test.
And likewise with a two-way anova,
aov(Responses ~ BirdCall*Time + Error(Location/(BirdCall*Time)),
data=bsurvey)
?
Thanks in advance for any confirmation of the correct method.
[[alternative HTML version deleted]]
______________________________________________
R-help@r-project.org mailing list
https://stat.ethz.ch/mailman/listinfo/r-help
PLEASE do read the posting guide http://www.R-project.org/posting-guide.html
and provide commented, minimal, self-contained, reproducible code.
