Newsom 1

USP 656 Multilevel Regression

Winter 2006
bhjjfjjjjjjjjfjfk (ie! 0 mg! 5 mg! "n# $0 mg% &owever! if the rese"r'her w"nts to m"ke inferen'es
be(on# the )"rti'ul"r v"lues of the 'gh#f#gfg#fgsfgs#fgs#fg#ffffffffffkkkkk#fgh#ghble use# in the stu#(!
" r"n#om effe'ts mo#el is use# * 'ommon e+"m)le woul# be the use of )ubli' "rt works
re)resenting low! mo#er"te! "n# 'vb'b'b'vb'vb
Fixed and Random Coefficients in Multilevel Regression
,he r"n#om vs fi+e# #hg#fg#fg#fg#fg#fg#fg#fk"n# effe'ts is im)ort"nt in multilevel regression -n
multilevel regression mo#els! both level.$ 'v'vb'vbv'b to v"r( r"n#oml( "'ross grou)s /e'"use of
the "ssum)tions "bout their error #istributions! we '"ll their v"ri"n'es! oo=var(U0j) "n# 11=var(U1j),
0r"n#om 'oeffi'ients1
,his me"ns th"t we '"n think "bout 0j "n# 1j "s "kin to the r"n#om variables - #es'ribe#
"bove -nste"# of "ttem)ting to gener"li2e be(on# the )"rti'ul"r v"lues of the in#e)en#ent v"ri"ble!
we "re "ttem)ting to gener"li2e be(on# the )"rti'ul"r grou)s in the stu#( 3or inst"n'e! we m"( h"ve
50 'om)"nies! but we wish to gener"li2e to " l"rger universe of 'om)"nies when we e+"mine the
me"ns (inter'e)ts% or the +.( rel"tionshi) (slo)es%
-n &4M! v"ri"n'es for the inter'e)ts "n# slo)es "re estim"te# b( #ef"ult ,h"t is! when (ou
first 'onstru't " mo#el! U0j "n# U1j "re estim"te# b( #ef"ult &owever! the rese"r'her h"s the 'hoi'e
of setting U0j "n# U1j to be 2ero /( setting them to 2ero! we "re testing " mo#el in whi'h we "ssume
0j "n# 1j #o not v"r( r"n#oml( "'ross grou)s -n f"'t! their v"ri"n'e is "ssume# to be 2ero! so the(
"re "ssume# to be 'onst"nt or 0nonv"r(ing1 "'ross grou)s 3or e+"m)le! fi+e#! nonv"r(ing inter'e)ts
woul# im)l( the grou) "ver"ge for the #e)en#ent v"ri"ble is "ssume# to be e5u"l in e"'h grou)
*lthough we t()i'"ll( refer to this 'onstr"int "s 0fi+ing the inter'e)ts1 or 0fi+ing the slo)es!1 the term is
somewh"t loosel( "))lie#! be'"use we "re re"ll( "ssuming the( "re fi+e# and nonvarying
Intercept Model and Random Effects ANOVA
-n the sim)lest &4M mo#el! there "re no )re#i'tors We h"ve one level.$ e5u"tion "n# one level.2
e5u"tion6
0 ij j ij
Y R = +
where Y is the #e)en#ent v"ri"ble! 0j is the inter'e)t! "n# Rij is the resi#u"l or error 0j '"n be
thought of "s the me"n of e"'h grou) ,he level.2 e5u"tion "lso h"s no )re#i'tors in it7s sim)lest
form6
0 00 0 j j
U = +
where
0j
is the #e)en#ent v"ri"ble!
00
is the level.2 inter'e)t! "n# U
0j
is the level.2 error -n this
e5u"tion! 00 re)resents the gr"n# me"n or the me"n of the inter'e)ts U0j re)resents the #evi"tion of
e"'h me"n from the gr"n# me"n When the "ver"ge #evi"tion is l"rge! there "re l"rge grou)
#ifferen'es
-t '"n be shown th"t this is re"ll( the *N89* mo#el *N89* e+"mines the #evi"tions of grou) me"ns
from the gr"n# me"n /e'"use we "ssume th"t the grou) me"ns! re)resente# b( 0j "n#! thus! their
#evi"tions "re r"n#oml( v"r(ing! this mo#el is e5uiv"lent to the random effects *N89* mo#el We
'"n rewrite our two &4M e5u"tions "s " single e5u"tion! if we )lug in the level.2 terms into level.$
e5u"tion
00 0 ij j ij
Y U R = + +
-n this e5u"tion! 00 re)resents the gr"n# me"n! the U0j term re)resents the #evi"tions from the me"n
(or me"n #ifferen'es or tre"tment effe't! if (ou woul# like%! "n# Rij re)resents the within.grou) error or
v"ri"tion -n more m"them"ti'"ll( oriente# *N89* te+tbooks! (ou will see "n *N89* mo#el written
something like6
Newsom 2
USP 656 Multilevel Regression
Winter 2006
ij j ij
Y e = + +
-f the mo#el is "n r"n#om effe'ts *N89*! j is "ssume# to h"ve " r"n#om #istribution -n multilevel
regression! be'"use we "ssume U0j to v"r( r"n#oml(! the sim)le &4M mo#el with no level.$ or level.2
)re#i'tors is e5uiv"lent to the r"n#om effe'ts *N89* mo#el
When we "## )re#i'tors to the level.$ e5u"tion! the( "re 'ov"ri"tes "n# the mo#el be'omes "
r"n#om effe'ts *N:89* in whi'h the me"ns "re "#juste# for the 'ov"ri"te
Newsom 3
USP 656 Multilevel Regression
Winter 2006
Summary a!le
Random vs"
Fixed
#efinition Example $se in Multilevel
Regression
Varia!les Random variable6 ($% is "ssume# to be me"sure# with
me"surement error ,he s'ores "re " fun'tion of " true
s'ore "n# r"n#om error; (2% the v"lues 'ome from "n#
"re inten#e# to gener"li2e to " mu'h l"rger )o)ul"tion
of )ossible v"lues with " 'ert"in )rob"bilit( #istribution
(eg! norm"l #istribution%; (<% the number of v"lues in
the stu#( is sm"ll rel"tive to the v"lues of the v"ri"ble "s
it "))e"rs in the )o)ul"tion it is #r"wn from Fixed
variable6 ($% "ssume# to be me"sure# without
me"surement error; (2% #esire# gener"li2"tion to
)o)ul"tion or other stu#ies is to the s"me v"lues; (<%
the v"ri"ble use# in the stu#( 'ont"ins "ll or most of the
v"ri"ble7s v"lues in the )o)ul"tion
-t is im)ort"nt to #istinguish between " v"ri"ble th"t is
varying "n# " v"ri"ble th"t is random * fi+e# v"ri"ble
'"n h"ve #ifferent v"lues! it is not ne'ess"ril( inv"ri"nt
(e5u"l% "'ross grou)s
Random variable:
)hotogr")hs
re)resenting in#ivi#u"ls
with #iffering levels of
"ttr"'tiveness
m"ni)ul"te# in "n
e+)eriment! 'ensus
tr"'ks
Fixed variable:
gen#er! r"'e! or
intervention vs 'ontrol
grou)
Pre#i'tor
v"ri"bles in M4R
gener"ll(
"ssume# to be
fi+e#
Effects Random effect6 ($% #ifferent st"tisti'"l mo#el of
regression or *N89* mo#el whi'h "ssumes th"t "n
in#e)en#ent v"ri"ble is r"n#om; (2% gener"ll( use# if
the levels of the in#e)en#ent v"ri"ble "re thought to be
" sm"ll subset of the )ossible v"lues whi'h one wishes
to gener"li2e to; (<% will )rob"bl( )ro#u'e l"rger
st"n#"r# errors (less )owerful% Fixed effect6 ($%
st"tisti'"l mo#el t()i'"ll( use# in regression "n# *N89*
"ssuming in#e)en#ent v"ri"ble is fi+e#; (2%
gener"li2"tion of the results "))l( to simil"r v"lues of
in#e)en#ent v"ri"ble in the )o)ul"tion or in other
stu#ies; (<% will )rob"bl( )ro#u'e sm"ller st"n#"r#
errors (more )owerful%
Random effect6
r"n#om effe'ts *N89*!
r"n#om effe'ts
regression
Fixed effect6 fi+e#
effe'ts *N89*! fi+e#
effe'ts regression
-nter'e)t onl(
mo#els in M4R
"re e5uiv"lent to
r"n#om effe'ts
*N89* or
*N:89*
Coefficients Random coefficient: term "))lies onl( to M4R
"n"l(ses in whi'h inter'e)ts! slo)es! "n# v"ri"n'es '"n
be "ssume# to be r"n#om M4R "n"l(ses most
t()i'"ll( "ssume r"n#om 'oeffi'ients 8ne '"n
'on'e)tu"li2e the 'oeffi'ients obt"ine# from the level.$
regressions "s " t()e of r"n#om v"ri"ble whi'h 'omes
from "n# gener"li2es to " #istribution of )ossible v"lues
=rou)s "re 'on'eive# of "s " subset of the )ossible
grou)s
Fixed coefficient: " 'oeffi'ient '"n be fi+e# to be non.
v"r(ing (inv"ri"nt% "'ross grou)s b( setting its between
grou) v"ri"n'e to 2ero
R"n#om 'oeffi'ients must be v"ri"ble "'ross grou)s
:on'e)tu"ll(! fi+e# 'oeffi'ients m"( be inv"ri"nt or
v"r(ing "'ross grou)s
Random coefficient6
the level.2 )re#i'tor!
"ver"ge in'ome! is
use# to )re#i't s'hool
)erform"n'e in e"'h
s'hool -nter'e)t
v"lues for s'hool
)erform"n'e "re
"ssume# to be "
s"m)le of the inter'e)ts
from " l"rger )o)ul"tion
of s'hools
Fixed coefficient6
slo)es or inter'e)ts
'onstr"ine# to be e5u"l
over #ifferent s'hools
/oth use# in
M4R Slo)es "n#
inter'e)t v"lues
'"n be
'onsi#ere# to be
fi+e# or r"n#om!
#e)en#ing on
rese"r'hers
"ssum)tions "n#
how the mo#el is
s)e'ifie# ,he
v"ri"n'e of the
slo)es or
inter'e)ts "re
'onsi#ere#
r"n#om
'oeffi'ients