Diagnostics: R [Study As­sess­ment]

posted by Helmut Homepage – Vienna, Austria, 2016-05-23 03:22 (2857 d 02:24 ago) – Posting: # 16349
Views: 29,312

Hi zizou,

❝ DF, p-values, CIs not reported in lmer for the reasons stated there.


Yep, we know. :-D

❝ library lmerTest can be used,


THX; I forgot!

❝ I think there is a little bug in df visualization - rounded on 3 decimal places, visible on 5 (actually I have not the latest version of R).


Same in R 3.2.5 and lmerTest 2.0-30.

summary(muddle)$coefficients["(Intercept)","df"]


❝  # [1] 9.195607


Or easier:
summary(muddle)$coefficients[, "df"]
(Intercept)   log(Dose)
   9.195607    5.896110


summary(muddle, digits=8) doesn’t help. Or as Carl Witthoft wrote: IMHO the proper use of summary is indicated by its name: a way to get a quick-look at your data.

❝ For calculation of 90% CIs: (below is only example for Intercept lower and upper limit)

alpha=0.05

❝ summary(muddle)$coefficients["(Intercept)","Estimate"] - qt(1-alpha,summary(muddle)$coefficients["(Intercept)","df"]) * summary(muddle)$coefficients["(Intercept)","Std. Error"]


Perfect! Little bit shorter:
summary(muddle)$coefficients[1, 1] +c(-1,+1)*qt(1-0.05, summary(muddle)$coefficients[1, 3]) * summary(muddle)$coefficients[1, 2]
[1] 1.496758 2.386013
summary(muddle)$coefficients[2, 1] +c(-1,+1)*qt(1-0.05, summary(muddle)$coefficients[2, 3]) * summary(muddle)$coefficients[2, 2]
[1] 0.6695768 0.8539044


So what do we have? A fair agreement across software – except SAS

Software               Estimate       90% CI       width
NCSS               B0 1.9414    1.4849    2.3978    0.9129
                   B1 0.7617    0.6659    0.8576    0.1917
Phoenix/WinNonlin  B0 1.9413858 1.4967592 2.3860125 0.8893
                   B1 0.7617406 0.6695783 0.8539029 0.1843
R lmerTest         B0 1.9413858 1.4967584 2.3860132 0.8893
                   B1 0.7617406 0.6695768 0.8539044 0.1843
SAS (code?)        B0 1.94      1.54      2.35      0.8100
                   B1 0.7615    0.679     0.844     0.1650

Since I trust most in Phoenix and R; given the width of the CI: Is NCSS too conservative and SAS liberal? Unfortunately we don’t have Smith’s code. Would be great if one of our SASians could jump in.

Dif-tor heh smusma 🖖🏼 Довге життя Україна! [image]
Helmut Schütz
[image]

The quality of responses received is directly proportional to the quality of the question asked. 🚮
Science Quotes

Complete thread:

UA Flag
Activity
 Admin contact
22,940 posts in 4,812 threads, 1,639 registered users;
39 visitors (0 registered, 39 guests [including 8 identified bots]).
Forum time: 04:47 CET (Europe/Vienna)

Those people who think they know everything
are a great annoyance to those of us who do.    Isaac Asimov

The Bioequivalence and Bioavailability Forum is hosted by
BEBAC Ing. Helmut Schütz
HTML5