GLM Shows Significance But Decision Tree Unable to Predict?

[Vik_R]

I'm trying to predict brand preference based on a number of attributes.

General Linear Model shows that all but one attribute has a significance of < 0.05, so one might expect I'd have a good ability to predict brand selection.

But running a Decision Tree shows that the correct brand is predicted only 25.7% of the the time.

What am I missing?

Thanks very much in advance to all for any info!

* Generalized Linear Models.
GENLIN BRAND_ID WITH Attribute_1 Attribute_2 Attribute_3 Attribute_4 Attribute_5 Attribute_6 
Attribute_7 Attribute_8 Attribute_9
/MODEL Attribute_1 Attribute_2 Attribute_3 Attribute_4 Attribute_5 Attribute_6 Attribute_7 
Attribute_8 Attribute_9 INTERCEPT=YES
DISTRIBUTION=POISSON LINK=LOG
/CRITERIA METHOD=FISHER(1) SCALE=1 COVB=MODEL MAXITERATIONS=100 MAXSTEPHALVING=5 
PCONVERGE=1E-006(ABSOLUTE) SINGULAR=1E-012 ANALYSISTYPE=3(WALD) CILEVEL=95 CITYPE=WALD 
LIKELIHOOD=FULL
/MISSING CLASSMISSING=EXCLUDE
/PRINT CPS DESCRIPTIVES MODELINFO FIT SUMMARY SOLUTION.

* Decision Tree.
TREE BRAND_ID [n] BY Attribute_1 [s] Attribute_2 [s] Attribute_3 [s] Attribute_4 [s] Attribute_5 
[s] Attribute_6 [s] Attribute_7 [s] Attribute_8 [s] Attribute_9 [s] 
/TREE DISPLAY=TOPDOWN NODES=STATISTICS BRANCHSTATISTICS=YES NODEDEFS=YES SCALE=AUTO
/DEPCATEGORIES USEVALUES=VALID
/PRINT MODELSUMMARY CLASSIFICATION RISK
/METHOD TYPE=CRT MAXSURROGATES=AUTO PRUNE=SE(1)
/GROWTHLIMIT MAXDEPTH=AUTO MINPARENTSIZE=100 MINCHILDSIZE=50
/VALIDATION TYPE=NONE OUTPUT=BOTHSAMPLES
/CRT IMPURITY=GINI MINIMPROVEMENT=0.0001
/COSTS EQUAL
/PRIORS FROMDATA ADJUST=NO.