%PDF-1.5
%
189 0 obj
<>
endobj
225 0 obj
<>/Font<>>>/Fields[]>>
endobj
247 0 obj
<>stream
Modeling categorical outcomes with random effects is a major use of the GLIMMIX procedure. Building, evaluating, and using the resulting model for inference, prediction, or both requires many considerations. This paper, written for expert users of SAS® statistical procedures, illustrates the nuances of the process with two examples: modeling a binary response using random effects and correlated errors, and modeling a multinomial response with random effects. In addition, the paper provides answers to common questions that are received by SAS Technical Support concerning these analyses with PROC GLIMMIX. These questions cover working with events and trials data, handling bias issues in a logistic model, and overcoming convergence problems. <br/><br/>Kathleen Kiernan, SAS
Session 2179
en
jeff.foxx@sas.com
Modeling categorical outcomes with random effects is a major use of the GLIMMIX procedure. Building, evaluating, and using the resulting model for inference, prediction, or both requires many considerations. This paper, written for expert users of SAS® statistical procedures, illustrates the nuances of the process with two examples: modeling a binary response using random effects and correlated errors, and modeling a multinomial response with random effects. In addition, the paper provides answers to common questions that are received by SAS Technical Support concerning these analyses with PROC GLIMMIX. These questions cover working with events and trials data, handling bias issues in a logistic model, and overcoming convergence problems. <br/><br/>Kathleen Kiernan, SAS
14.2
SAS/STAT software
PROC GLIMMIX, GLIMMIX procedure, categorical outcomes
PScript5.dll Version 5.2.2
2018-04-02T16:03:34.000-04:00
2018-04-02T16:03:34.000-04:00
2018-01-30T16:01:11.000-05:00
2018-04-03T15:45:53.496-04:00
3147ad82bfba7deb002b3cf9709bc02a36f716d1
576018
application/pdf
2018-04-03T15:45:53.262-04:00
SAS Institute Inc. (author: Kathleen Kiernan)
Insights into Using the GLIMMIX Procedure to Model Categorical Outcomes with Random Effects
uuid:6ceceab5-1b72-42c9-9e99-dd8a39e7e752
uuid:87302eff-1753-41cc-af6a-34d47c08b6b0
thirdparty
Acrobat Distiller 11.0 (Windows)
support:sgf-papers
year:2018
software:STAT
support:sgf-papers/session-type/breakout
support:sgf-papers/skill-level/intermediate
support:sgf-papers/topic/analytics/statistics
support:customer-roles/statistician
endstream
endobj
181 0 obj
<>
endobj
183 0 obj
<>
endobj
184 0 obj
<>
endobj
185 0 obj
<>
endobj
186 0 obj
<>
endobj
86 0 obj
<>/Font<>/ProcSet[/PDF/Text]>>/Rotate 0/Type/Page>>
endobj
88 0 obj
<>/ExtGState<>/Font<>/ProcSet[/PDF/Text]>>/Rotate 0/Type/Page>>
endobj
94 0 obj
<>/ExtGState<>/Font<>/ProcSet[/PDF/Text]>>/Rotate 0/Type/Page>>
endobj
130 0 obj
[131 0 R 132 0 R 133 0 R 134 0 R 135 0 R 136 0 R 137 0 R]
endobj
95 0 obj
<>stream
HW]o[9+z|0;M٧ mv8%)Jⵝb/ͽEQ9uV_bkʨ/JOZnv(
rZO%*bͣ&/W6CpU҅7[pokpNh䎶voᗵa2y
8W_ø|
W7!s7c_?e`6v^T?13O<.Ytj4wȧk/am5_ݯ0J607Gp3DʂzѦ&q!՛"BKr4.ipuv+]öHm262il:[DåP̹ͣQ#.{Pg<' 9Zq`GC@0]w'4zt[~~Nj(.!E$;IMn1Y߿SfVx|0N
BDϼZwUV%˨}$G[WLRKyiiΧ%xߗDnp]]/Wf?"zkmcg'>òv
e1- f,V4x3u5F0rq2P
t&zжZrD`4r?lBnJ"m6z{qOhw`Ĩ_RiIԾ')L>#-op{vNjewx}?/d,fٕd,>O / *`[7Lf]`t':__Ia0YSE[0Tpw˵E LC 9Qe.
;=cK̋4h\-f:t}+f۷cCߖUzAvrĭPMkP3Pdʩ-RMT%jxk1j[Q4Okbz9938L@zT4Mb]20RvcEͰN:0(sb+b|%ٕܐSgH5g,͉!*^Dꋧ6]-
NACPH| <>vlI}~}m<:fA8܇eׇYn5ZQ oFbLSr'M9ĭAY7W2Fc1̧x1tkn.WZ
NǞX>mY2I6
S-K$Qad+7J']>S f.":qR3~#Z k'4}дVpE>NCu
@җ#ĦIu+4|uB(kIHٹ AEEGϦp9`CN
oZw+i(kusU5pvM4a1d*'ky5vA;+].u=X
T,1wѓ!|Ӧs+U&3oUvlQj;(Ϧ^f<;9-rw"ZiYzk2`02ToDJS#2bܾUv H>~Wݲg 3;]͟j
!cAoU r"ڇ^j i4rUn\:W%S:Q1
'z4 8֑
|=9Rxw}7yxܰ$Of%G'Q3P)hJ7ė%P˵TX"N 8,yCMm\O7Ed
)LQ87Z%ьp8'Z%)O-j}
6Kiۄ4d* av2q?(31AAmlt%ZyqY1טawdva㥴>x"lGAb
Jf4=|jtTX?EG!rZY(;ZYՂdYz ަ_o`-<:LFwh3"dU"
cqd Ijw`{BtҒ"dKRmz{~Nô?^kX0biawJ97ƀ"̡9)j@8Z}ER
!N,28tküxa4s,lza=KrX:06ya8V(=fNPDZ`*sKrYygֹ)25s=%*,gaZlBKƝM}cD;qw`:4% Zj#4_k~fwL7Al!3U-wG跐<%RQ[dzřvd>kwici3flr3zv+0P4>8f9=gDjꀄ=DLW(4_Fq6:BݨWũ=*_@u;T>:GcNVi[8L|4DBՎmӾ!PCivn]7gڲGԁ0d=A$7?)Gei$ U̳:S=|t駩$Co,f=" ڟ)}PBCz_#1=%_R")x]jl~^ġDSDq"HfhӱIB0*h^YTwH<#iM@gāC̲)4 J1RcWz_em$ï¥օ,1 qo:W+щl`O?ZuNfa"^wXVgaNFXp.eLү{hy)$զNђ!`-y:Z)uiKq*ࠧd{QQxl|`mïa Q\9
h_`,KAֽg,J(̂
4R.s\ZoJ&T1¶lDqq1xȊp?N4A-:
ы$jw~kMM]6z=*]b!@p.E(%<`%8#vE
PAQT;rGvo. F9
k` 7$z:P;wܫh 4^%Ҷ
SԀ@*D5tMF$m%cҠ(& AXE#sjゃffΝ!uqƏgHocLtLd*uq@k/Sס5E>'@u&4Ȧa'(c:lPcFŻ#Y.4|@&FB抓_ۖRӱZ&zفr*1y+{or\Sur6du_Hʤ1bʳZϬbʤʁBG/Cg$8t '\FѠNKL {nTWi!r)lai(VPLnTmyo`$suo6Z&ak#23*v~P7]S85
Ui*H^l/~EJa`F\&qjunC_!(-L(o0Dz֗e.0Yu
H!rS1Ziz$m
;5&ē̚rk6Eo-)ǢZ匬Anꕗsg),Kٯ,|1\B<r-t\6U ;UuԄq~oPVv}\קS<5s+ٴplj{V~yUKOmQ5*+Gk^vbmnCay4k}",gz+S0V뜯,羇G؇JNaJXUBce