2

I want to use a MANOVA test in R to check for a statistical difference between a vector of means ("test") and a vector (of equal length) of '1's ("random"). My data is as below:

    ID      openwater    closedshrubland      barren    cropnatural         crop   decidbroad     mixed
  test      0.8435707           1.037015     0.90612      0.8724474    0.9837284     1.035505   1.05059
random              1                  1           1              1            1            1         1

When I run the following code:

manovares1 <- manova(cbind(openwater,closedshrubland,barren,cropnatural,crop,decidbroad,mixed) ~ as.factor(ID),data=test)

The print out gives me the following message:

Error in print.aov(list(coefficients = c(1, -0.156429334, 1, 0.0370145719999997,  : 
  length of 'dimnames' [1] not equal to array extent

And the summary(manovares1) gives:

              Df
as.factor(ID)  1**

Any ideas what I am doing wrong? (A similar question was asked here, but the answer doesn't seem to help in my situation.

Thanks for any help you can offer!

Heather
  • 45
  • 1
  • 5
  • I have no idea how you expect the MANOVA to estimate variances if each of your groups contains exactly one value vector. – Roland Aug 13 '14 at 08:15

1 Answers1

0

I'm not sure what's going on with your database, but I bet that's the problem.

What exactly are your IDs, is it the openwater, etc.? In this case, this is what's wrong, cause every group in the factor has exactly one case. No program will give you better results when looking at variance. Since there is none.

Orsolya M
  • 38
  • 6