Dear Laura,<br><br>this model is most certainly overfit. for a binomial model the rule of thumb for the absolute upper limit for the number of paramters in the model is min(outcome A, outcome B) / 10, which in your case is 12/ 10 ~ 1. you're using 2 parameters for fixed effects (and 1 for a random intercept).<br>
<br>Most of your subjects will have 0 deaccented syllables, making it hard to impossible for the model to estimate the effects of info. I am still a bit puzzled by the output though. did the fitting process not through a warning (I am leaving town tomorrow, so i did not have time to load your data and see for myself).<br>
<br>Florian<br><br><br><div class="gmail_quote">On Mon, Dec 22, 2008 at 4:57 AM, Laura de Ruiter <span dir="ltr"><<a href="mailto:Laura.deRuiter@mpi.nl">Laura.deRuiter@mpi.nl</a>></span> wrote:<br><blockquote class="gmail_quote" style="border-left: 1px solid rgb(204, 204, 204); margin: 0pt 0pt 0pt 0.8ex; padding-left: 1ex;">
Dear R-users and -experts,<br>
<br>
I am performing a rather simple analysis on a small data set (pasted below this email) and keep getting a to me inexplicable result. Perhaps I am missing something here - it would be great if someone could point out to me what I am doing wrong.<br>
<br>
I want to test whether the factor "Info" (which has three levels: "new", "given", "accessible") is a significant predictor for the binary variable "DeaccYN". The random factor is "Subject". The distribution of the data looks as follows:<br>
<br>
-----------------------------------------------------------------------------<br>
xtabs (~ DeaccYN + Info, data = dat)<br>
Info<br>
DeaccYN given new accessible<br>
no 25 42 21<br>
yes 11 0 1<br>
------------------------------------------------------------------------------<br>
<br>
This is the model:<br>
<br>
----------------------------------------------------------------------------------------------------------<br>
deacc.lmer = lmer (DeaccYN ~ Info + (1|Subject), data = dat, family = "binomial")<br>
-----------------------------------------------------------------------------------------------------------------<br>
<br>
However, given the distribution above, this outcome seems rather weird to me:<br>
<br>
---------------------------------------------------------------------------------------------------------<br>
summary (deacc.lmer)<br>
Generalized linear mixed model fit using Laplace<br>
Formula: DeaccYN ~ Info + (1 | Subject)<br>
Data: dat<br>
Family: binomial(logit link)<br>
AIC BIC logLik deviance<br>
60.4 70.82 -26.2 52.4<br>
Random effects:<br>
Groups Name Variance Std.Dev.<br>
Subject (Intercept) 0.18797 0.43356<br>
number of obs: 100, groups: Subject, 21<br>
<br>
Estimated scale (compare to 1 ) 0.7316067<br>
<br>
Fixed effects:<br>
Estimate Std. Error z value Pr(>|z|)<br>
(Intercept) -0.8635 0.3795 -2.2754 0.0229 *<br>
Infonew -18.7451 2764.2445 -0.0068 0.9946<br>
Infoaccessible -2.2496 1.1186 -2.0110 0.0443 *<br>
---<br>
Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1<br>
<br>
Correlation of Fixed Effects:<br>
(Intr) Infonw<br>
Infonew 0.000<br>
Infoaccssbl -0.315 0.000<br>
----------------------------------------------------------------------------------------------------<br>
<br>
Why should the difference between 25/11 and 21/1 be significant, but the difference between 25/11 and 42/0 not? Very odd to me seems the standard error of 2764!<br>
First I was wondering whether the problem had sth. to do with the fact that I had read in an SPSS file, but I keep getting this result even when I first save the SPSS file as a tabdelim. text and read it into R then. I couldn't find anything strange going on with the Subject factor either (and I have also tried the model with "Item" as random factor, which didn't change anything).<br>
<br>
I was wondering: Is it perhaps a problem for the model that there are no cases in the DeaccYN == "yes" category for Info == "given"? And if this is the case, why?<br>
Am I overlooking something here?<br>
<br>
Any help with this would be very much appreciated!<br>
<br>
Thanks a lot in advance for your answers & happy holidays,<br>
<br>
Laura de Ruiter<br>
<br>
---------------------------------------<br>
The data set:<br>
<br>
dat<br>
Subject Info Deacc<br>
1 93 new no<br>
2 93 new no<br>
3 93 given yes<br>
4 93 new no<br>
5 93 given no<br>
6 94 new no<br>
7 94 new no<br>
8 94 given no<br>
9 94 new no<br>
10 94 given no<br>
11 94 given no<br>
12 94 accessible no<br>
13 95 given no<br>
14 95 new no<br>
15 95 accessible no<br>
16 95 new no<br>
17 95 accessible no<br>
18 95 new no<br>
19 97 new no<br>
20 98 new no<br>
21 98 new no<br>
22 98 given yes<br>
23 98 given no<br>
24 101 new no<br>
25 101 new no<br>
26 101 given yes<br>
27 101 accessible no<br>
28 101 new no<br>
29 101 given yes<br>
30 101 given no<br>
31 101 accessible no<br>
32 102 new no<br>
33 102 new no<br>
34 102 given no<br>
35 102 given no<br>
36 102 accessible no<br>
37 104 new no<br>
38 104 accessible no<br>
39 104 new no<br>
40 104 given yes<br>
41 104 new no<br>
42 104 given no<br>
43 104 accessible no<br>
44 107 new no<br>
45 107 given no<br>
46 107 accessible no<br>
47 107 new no<br>
48 109 new no<br>
49 109 given yes<br>
50 109 new no<br>
51 110 accessible no<br>
52 110 new no<br>
53 110 new no<br>
54 110 given no<br>
55 113 new no<br>
56 113 given yes<br>
57 113 new no<br>
58 113 given yes<br>
59 114 new no<br>
60 114 accessible no<br>
61 114 new no<br>
62 114 given no<br>
63 114 accessible no<br>
64 114 new no<br>
65 114 given yes<br>
66 114 accessible yes<br>
67 115 accessible no<br>
68 115 given no<br>
69 115 given no<br>
70 116 accessible no<br>
71 116 new no<br>
72 116 given no<br>
73 116 new no<br>
74 116 new no<br>
75 116 accessible no<br>
76 117 accessible no<br>
77 117 given no<br>
78 117 new no<br>
79 117 new no<br>
80 117 given no<br>
81 117 accessible no<br>
82 118 new no<br>
83 118 accessible no<br>
84 118 given no<br>
85 118 new no<br>
86 118 given no<br>
87 118 accessible no<br>
88 119 given no<br>
89 119 given yes<br>
90 119 given no<br>
91 119 accessible no<br>
92 121 given no<br>
93 121 given no<br>
94 131 new no<br>
95 131 new no<br>
96 131 new no<br>
97 131 new no<br>
98 132 given yes<br>
99 132 given no<br>
100 132 accessible no<br>
------------------------------------------------<br>
<br>
-- <br>
//////////////////////////////////////////////////////<br>
<br>
Laura de Ruiter<br>
PhD student<br>
Language Acquisition Group<br>
Max Planck Institute for Psycholinguistics<br>
P.O. Box 310<br>
6500 AH Nijmegen<br>
Netherlands<br>
<br>
Phone: 0031-24-3521 456<br>
Homepage: <a href="http://www.mpi.nl/Members/LauradeRuiter" target="_blank">http://www.mpi.nl/Members/LauradeRuiter</a><br>
<br>
///////////////////////////////////////////////////<br>
<br>
<br>
_______________________________________________<br>
R-lang mailing list<br>
<a href="mailto:R-lang@ling.ucsd.edu" target="_blank">R-lang@ling.ucsd.edu</a><br>
<a href="http://pidgin.ucsd.edu/mailman/listinfo/r-lang" target="_blank">http://pidgin.ucsd.edu/mailman/listinfo/r-lang</a><br>
</blockquote></div><br>