[R] 2x2x2 rm ANOVA, varying results

[Jonathan Reardon]

Hello,

I ran a 2x2x2 repeated measures ANOVA which turned out fine:
                                     Df                Sum Sq       Mean Sq   F value     Pr(>F)  Attend                        1                   0.5540      0.55402     7.0374    0.01079 *PercGrp                      1                   0.0058      0.00580     0.0737    0.78719  Pres                             1                   0.1794      0.17944     2.2794    0.13766  Attend:PercGrp         1                   0.0017      0.00172     0.0218    0.88324  Attend:Pres                 1                 0.0189      0.01894    0.2406      0.62598  PercGrp:Pres              1                  0.0534      0.05344    0.6789     0.41405  Attend:PercGrp:Pres  1                 0.0046     0.00464    0.0590      0.80912  Residuals                      48                3.7788  0.07872    
However, when I run the interactions alone (from the same dataset), I get a different set of results than what was originally shown e.g.
> anova(lm(main~Attend+PercGrp+Pres))Analysis of Variance Table
Response: main                  Df       Sum Sq    Mean Sq    F value      Pr(>F)   Attend     1        0.5540       0.55402    7.4682     0.008561 **PercGrp   1       0.0058        0.00580    0.0782     0.780849   Pres         1        0.1794        0.17944    2.4189     0.125942   Residuals 52     3.8575        0.07418  
I also get different results when i run the interactions alone too. Curious to know why this is.
Thanks,Jon




 		 	   		  
	[[alternative HTML version deleted]]