FITS in Minitab

#1
Hi,

I just carried out an ANOVA on my data where I used blocks. There is 8 blocks and 4 treatment and I laid out my data as below.

Thing I need to know the FITS to find out my contrast sum of squares. But I dont understand how Minitab computed FITS here. Previously when I didn't use blocks, all resp corresponding to a treatment would have fixed FITS but ehre its all over the place. For my contrasts, say I want to compare treatment 1 against 2, I dunno what to pick.

My data set is provided below. This is only first 6 of the dataset just to show the problem, not the whole thing. I used GLM in Minitab and set resp as RESPONSE and 'block' and 'tmt' in MODEL and left 'RANDOM FACTOR' blank.

There is no negative signs, its just padded with dashes to make it readable.

Hope someone can tell me how minitab worked out hte FITS below.

ob-block-tmt---resp-----res1-------fits1
1---1-----1----331-----23.34-----307.656
2---1-----2----317-----58.91-----375.906
3---1-----3----479----104.09------374.906
4---1-----4----245-----68.531-----313.531
5---2-----1----402-----50.844-----351.156
6---2-----2----423-----3.594------419.406


Cheers
 
Last edited: