anova - P-value Troubles in R -
i have question regarding p-values. i've been comparing different linear models determine if 1 model better following function in r.
anova(model1,model2)
unfortunately, not calculate f or p-value. here example of anova summary did not give p-value
analysis of variance table model 1: influence ~ sortedsums[, combos2[1, a]] + sortedsums[, combos2[2,a]] model 2: influence ~ sortedsums[, b] res.df rss df sum of sq f pr(>f) 1 127 3090.9 2 128 2655.2 -1 435.74
for sake of symmetry, here anova summary did yield p-value.
analysis of variance table model 1: influence ~ sortedsums[, combos2[1, a]] + sortedsums[, combos2[2,a]] model 2: influence ~ sortedsums[, b] res.df rss df sum of sq f pr(>f) 1 127 3090.9 2 128 3157.6 -1 -66.652 2.7386 0.1004
do know why occurs?
not questions require code examples. don't deserve snarked @ being new, , i'm sorry people did. here answer:
the difference between 2 models not significant.
here can it:
- check make sure terms of 1 model object superset of terms of other. otherwise, default anova test invalid begin (you instead compare such non-nested models using aic, belongs in separate question). i'm curious see nested pair of models manages that non-significant, again, it's not necessary answering question.
- if checked, , models nested, , analysis doing manually, write p=1.0 in report , call day.
- if models nested, , above feels cheating, here's how th hard way. asking
anova
whether 1 variable differ makes significant contribution fit. take "larger" model ,summary(bar)
. p-value corresponding variable present inbar
missing infoo
p-value! , it's equal 1. , square of t-statistic f-value. - if models nested , analysis doing programmatically , absence of p-value breaks stuff elsewhere in script,
anova(foo,bar)[,5:6]
na
s instead of blanks... again, if doing programmatically have tried that.
good luck!
Comments
Post a Comment