1 One OVERVIEW T he Cross-Battery Assessment (XBA) approach (hereafter referred to as the XBA approach) was introduced by Flanagan and her colleagues in the late 1990s (Flanagan & McGrew, 1997; Flanagan, McGrew, & Ortiz, 2000; McGrew & Flanagan, 1998). The XBA approach provides practitioners with the means to make systematic, valid, and up-to-date in- terpretations of intelligence batteries and to augment them with other tests (e.g., academic ability tests) in a way that is consistent with the empirically supported Cattell-Horn-Carroll (CHC) theory of cognitive abilities. Moving beyond the boundaries of a single intelligence test kit by adopting the psy- chometrically and theoretically defensible XBA principles and procedures represents a significantly improved method of measuring cognitive abilities (Carroll, 1998; Kaufman, 2000). According to Carroll (1997), the CHC taxonomy of human cognitive abili- ties “appears to prescribe that individuals should be assessed with respect to the total range of abilities the theory specifies” (p. 129, emphasis added). However, because Carroll recognized that “any such prescription would of course create enormous problems,” he indicated that “[r]esearch is needed to spell out how the assessor can select what abilities need to be tested in particular cases” (p. 129). Flanagan and colleagues’ XBA approach was de- veloped specifically to “spell out” how practitioners can conduct assessments that approximate the total range of broad and narrow cognitive abilities more adequately than what is possible with a single intelligence battery. In a review of the XBA approach, Carroll (1998) stated that it “can be used to develop the most appropriate information about an individual in a given testing situ- ation” (p. xi). In Kaufman’s (2000) review of the XBA approach, he stated that the approach is based on sound assessment principles, adds theory to COPYRIGHTED MATERIAL
37
Embed
OVERVIEW T COPYRIGHTED MATERIAL · 2020. 2. 20. · practitioners with the means to make systematic, valid, and up- to- date in- ... 2 ESSENTIALS OF CROSS-BATTERY ASSESSMENT ... solve
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
1
One
OVERVIEW
The Cross- Battery Assessment (XBA) approach (hereafter referred to as the XBA approach) was introduced by Flanagan and her colleagues in the late 1990s (Flanagan & McGrew, 1997; Flanagan, McGrew,
& Ortiz, 2000; McGrew & Flanagan, 1998). The XBA approach provides practitioners with the means to make systematic, valid, and up- to- date in-terpretations of intelligence batteries and to augment them with other tests (e.g., academic ability tests) in a way that is consistent with the empirically supported Cattell- Horn- Carroll (CHC) theory of cognitive abilities. Moving beyond the boundaries of a single intelligence test kit by adopting the psy-chometrically and theoretically defensible XBA principles and procedures represents a signifi cantly improved method of measuring cognitive abilities (Carroll, 1998; Kaufman, 2000).
According to Carroll (1997), the CHC taxonomy of human cognitive abili-ties “appears to prescribe that individuals should be assessed with respect to the total range of abilities the theory specifi es” (p. 129, emphasis added). However, because Carroll recognized that “any such prescription would of course create enormous problems,” he indicated that “[r]esearch is needed to spell out how the assessor can select what abilities need to be tested in particular cases” (p. 129). Flanagan and colleagues’ XBA approach was de-veloped specifi cally to “spell out” how practitioners can conduct assessments that approximate the total range of broad and narrow cognitive abilities more adequately than what is possible with a single intelligence battery. In a review of the XBA approach, Carroll (1998) stated that it “can be used to develop the most appropriate information about an individual in a given testing situ-ation” (p. xi). In Kaufman’s (2000) review of the XBA approach, he stated that the approach is based on sound assessment principles, adds theory to
COPYRIG
HTED M
ATERIAL
2 ESSENTIALS OF CROSS-BATTERY ASSESSMENT
psychometrics, and improves the quality of the assessment and interpretation of cognitive abilities and processes.
Noteworthy is the fact that the “crossing” of batteries is not a new method of intellectual assessment. Neuropsychological assessment has long adopted the practice of crossing various standardized tests in an attempt to measure a broader range of brain functions than that offered by any single instrument (Lezak, 1976, 1995). Nevertheless, several problems with crossing batteries have plagued assessment- related fi elds for years. Many of these problems have been circumvented by Flanagan and colleagues’ XBA approach (see Rapid Reference 1.1 for examples).
Unlike the XBA model, the various so- called “cross- battery” techniques applied within the fi eld of neuropsychological assessment, for example, are not grounded in a systematic approach that is both psychometrically and the-oretically defensible. Thus, as Wilson (1992) cogently pointed out, the fi eld of neuropsychological assessment is in need of an approach that would guide practitioners through the selection of measures that would result in more specifi c and delineated patterns of function and dysfunction—an approach that provides more clinically useful information than one that is “wedded to the utilization of subscale scores and IQs” (p. 382). Indeed, all fi elds in-volved in the assessment of cognitive functioning have some need for an ap-proach that would aid practitioners in their attempt to “touch all of the major cognitive areas, with emphasis on those most suspect on the basis of history, observation, and on- going test fi ndings” (Wilson, 1992, p. 382). The XBA ap-proach represents a quantum leap in this direction. Recently, other researchers appear to be offering similar recommendations as those inherent in the XBA approach (e.g., Dehn, 2006; Fiorello & Hale, 2006). The defi nition of XBA as well as the rationale and foundations for and applications of this approach are depicted in Figure 1.1 and are described briefl y in the following sections.
DEFINITION
The XBA approach is a time- effi cient method of cognitive assessment that is grounded in CHC theory and research. It allows practitioners to reliably measure a wider range (or a more in- depth but selective range) of cogni-tive abilities / processes than that represented by a single intelligence battery. The XBA approach is based on three foundational sources of information or
Par
alle
l Nee
ds
in C
ogn
itiv
e A
sses
smen
t- R
elat
ed F
ield
s A
ddre
ssed
by
the
XB
A A
ppro
ach
Nee
d w
ith
in A
sses
smen
t- R
elat
ed F
ield
saN
eed
Ad
dre
ssed
by
the
XB
A A
pp
roac
h
Scho
ol P
sych
olog
y, C
linic
al P
sych
olog
y, an
d N
euro
psyc
holo
gy
have
lagg
ed in
the
dev
elop
men
t of c
once
ptua
l mod
els
of t
he
asse
ssm
ent o
f ind
ivid
uals
. The
re is
a n
eed
for
the
deve
lopm
ent
of c
onte
mpo
rary
mod
els.
The
XBA
app
roac
h pr
ovid
es a
con
tem
pora
ry c
once
ptua
l m
odel
for
mea
sure
men
t and
inte
rpre
tatio
n of
hum
an c
ogni
tive
abili
ties.
It is
likel
y th
at t
here
is a
nee
d fo
r ev
ents
ext
erna
l to
a fi e
ld o
f en
deav
or to
giv
e im
petu
s to
new
dev
elop
men
ts a
nd r
eal a
d-va
nces
in t
hat fi
eld
.
Car
roll
and
Hor
n’s
Flui
d- Cr
ysta
llized
the
oret
ical
mod
els
and
sys-
tem
atic
pro
gram
s of
res
earc
h in
cog
nitiv
e ps
ycho
logy
pro
vide
d th
e im
petu
s fo
r th
e X
BA a
ppro
ach
and
led
to t
he d
evel
opm
ent
of b
ette
r as
sess
men
t ins
trum
ents
and
pro
cedu
res.
The
re is
a n
eed
for
trul
y un
idim
ensio
nal a
sses
smen
t ins
tru-
men
ts fo
r ch
ildre
n an
d ad
ults
. With
out t
hem
, val
id in
terp
reta
-tio
ns o
f tes
t sco
res
are
prob
lem
atic
at b
est.
Man
y sc
ale
and
com
posi
te m
easu
res
on in
telli
genc
e ba
tter
ies
are
mix
ed, c
onta
inin
g ex
cess
rel
iabl
e va
rianc
e as
soci
ated
with
a
cons
truc
t irr
elev
ant t
o th
e on
e in
tend
ed fo
r m
easu
rem
ent a
nd
inte
rpre
tatio
n. T
he X
BA a
ppro
ach
ensu
res
that
ass
essm
ents
in
clud
e co
mpo
site
s or
clu
ster
s th
at a
re r
elat
ivel
y pu
re r
epre
-se
ntat
ions
of C
HC
bro
ad a
nd n
arro
w a
bilit
ies,
allo
win
g fo
r va
lid
mea
sure
men
t and
inte
rpre
tatio
n of
mul
tiple
uni
dim
ensio
nal
cons
truc
ts.
Rapid
Refer
ence
1.1
(con
tinue
d)
Nee
d w
ith
in A
sses
smen
t- R
elat
ed F
ield
saN
eed
Ad
dre
ssed
by
the
XB
A A
pp
roac
h
The
re is
a n
eed
to u
tiliz
e a
conc
eptu
al fr
amew
ork
to d
irect
any
ap
proa
ch to
ass
essm
ent.
Thi
s w
ould
aid
in b
oth
the
sele
ctio
n of
inst
rum
ents
and
met
hods
, and
in t
he in
terp
reta
tion
of te
st
fi ndi
ngs.
The
XBA
app
roac
h to
ass
essm
ent i
s ba
sed
on C
HC
the
ory.
Sinc
e th
is ap
proa
ch li
nks
all t
he m
ajor
inte
llige
nce
batt
erie
s (a
nd
a va
riety
of o
ther
cog
nitiv
e ab
ility
/ pro
cess
ing
and
acad
emic
te
sts)
to t
his
theo
ry, b
oth
sele
ctio
n of
test
s an
d in
terp
reta
tion
of te
st fi
ndin
gs a
re m
ade
easy
.
It is
nece
ssar
y th
at t
he c
once
ptua
l fra
mew
ork
or m
odel
und
er-
lyin
g as
sess
men
t inc
orpo
rate
s va
rious
asp
ects
of n
euro
psyc
ho-
logi
cal a
nd c
ogni
tive
abili
ty fu
nctio
n th
at c
an b
e de
scrib
ed in
te
rms
of c
onst
ruct
s th
at a
re r
ecog
nize
d in
the
neu
rops
ycho
logi
-ca
l and
cog
nitiv
e ps
ycho
logy
lite
ratu
re.
The
XBA
app
roac
h in
corp
orat
es v
ario
us a
spec
ts o
f neu
rops
y-ch
olog
ical
and
cog
nitiv
e ab
ility
func
tions
tha
t are
des
crib
ed in
te
rms
of c
onst
ruct
s th
at a
re r
ecog
nize
d in
the
rel
ated
lite
ra-
ture
.
The
re is
a n
eed
to a
dopt
a c
once
ptua
l fra
mew
ork
that
allo
ws
for
the
mea
sure
men
t of t
he fu
ll ra
nge
of b
ehav
iora
l fun
ctio
ns
subs
erve
d by
the
bra
in. U
nfor
tuna
tely
, in
neur
opsy
chol
ogic
al
asse
ssm
ent t
here
is n
o in
clus
ive
set o
f mea
sure
s th
at is
sta
n-da
rdiz
ed o
n a
singl
e no
rmat
ive
popu
latio
n.
The
XBA
app
roac
h al
low
s fo
r th
e m
easu
rem
ent o
f a w
ide
rang
e of
bro
ad a
nd n
arro
w c
ogni
tive
abili
ties /
proc
esse
s sp
eci-
fi ed
in c
onte
mpo
rary
CH
C t
heor
y. A
lthou
gh a
n X
BA n
orm
gr
oup
does
not
exi
st, t
he c
hara
cter
istic
s of
the
nor
mal
pro
b-ab
ility
cur
ve a
nd s
ound
psy
chom
etric
prin
cipl
es a
re u
sed
to
inte
rpre
t XBA
dat
a ef
fect
ivel
y.
Beca
use
ther
e ar
e no
tru
ly u
nidi
men
siona
l mea
sure
s in
psy
cho-
logi
cal a
sses
smen
t, th
ere
is a
need
to s
elec
t sub
test
s fr
om s
tan-
dard
ized
inst
rum
ents
tha
t app
ear
to r
efl e
ct t
he n
euro
cogn
itive
fu
nctio
n of
inte
rest
. In
neur
opsy
chol
ogic
al a
sses
smen
t, th
e ai
m,
ther
efor
e, is
to s
elec
t tho
se m
easu
res
that
, on
the
basis
of c
are-
ful t
ask
anal
ysis
, app
ear
mai
nly
to t
ap a
giv
en c
onst
ruct
.
The
XBA
app
roac
h is
defi n
ed b
y a
CH
C c
lass
ifi ca
tion
syst
em.
Subt
ests
from
the
maj
or in
telli
genc
e ba
tter
ies
(and
var
ious
ot
her
inst
rum
ents
) w
ere
clas
sifi e
d em
piric
ally
as
mea
sure
s of
br
oad
and
narr
ow C
HC
con
stru
cts.
Use
of t
hese
cla
ssifi
catio
ns
allo
ws
prac
titio
ners
to b
e re
ason
ably
con
fi den
t tha
t a g
iven
test
ta
ps a
giv
en c
onst
ruct
.
It is
clea
r th
at a
n ec
lect
ic a
ppro
ach
is ne
eded
in t
he s
elec
tion
of
mea
sure
s, pr
efer
ably
sub
test
s ra
ther
than
the
omni
bus
IQs,
in
orde
r to
gai
n m
ore
spec
ifi ci
ty in
the
del
inea
tion
of p
atte
rns
of
func
tion
and
dysf
unct
ion.
The
XBA
app
roac
h en
sure
s th
at t
wo
or m
ore
rela
tivel
y pu
re,
but q
ualit
ativ
ely
diffe
rent
, ind
icat
ors
of e
ach
broa
d co
gniti
ve
abili
ty / p
roce
ss a
re r
epre
sent
ed in
a c
ompl
ete
asse
ssm
ent o
f th
at c
onst
ruct
. Tw
o or
mor
e qu
alita
tivel
y si
mila
r in
dica
tors
are
ne
cess
ary
to m
ake
infe
renc
es a
bout
spe
cifi c
or
narr
ow C
HC
ab
ilitie
s. T
his
proc
ess
is ec
lect
ic in
its
sele
ctio
n of
mea
sure
s, bu
t at
tem
pts
to r
epre
sent
all
broa
d an
d na
rrow
abi
litie
s / pr
oces
ses
by u
sing
a su
bset
of m
easu
res
from
onl
y tw
o ba
tter
ies
(tha
t w
ere
norm
ed w
ithin
a fe
w y
ears
of o
ne a
noth
er).
The
re is
a n
eed
to s
olve
the
pot
entia
l pro
blem
s th
at c
an a
rise
from
cro
ssin
g no
rmat
ive
grou
ps a
s w
ell a
s se
ts o
f mea
sure
s th
at
vary
in r
elia
bilit
y.
In t
he X
BA a
ppro
ach,
one
can
typ
ical
ly a
chie
ve b
asel
ine
data
in
cog
nitiv
e fu
nctio
ning
acr
oss
seve
n or
eig
ht C
HC
bro
ad a
bil-
ities
/ pro
cess
es t
hrou
gh t
he u
se o
f onl
y tw
o w
ell- s
tand
ardi
zed
batt
erie
s, w
hich
min
imiz
es t
he e
ffect
s of
err
or d
ue to
nor
min
g di
ffere
nces
. Also
, sin
ce in
terp
reta
tion
of b
oth
broa
d an
d na
r-ro
w C
HC
abi
litie
s / pr
oces
ses
is m
ade
at t
he c
lust
er (
rath
er t
han
subt
est)
leve
l, iss
ues
rela
ted
to lo
w r
elia
bilit
y ar
e le
ss p
robl
em-
atic
in t
his
appr
oach
. Fin
ally
, bec
ause
con
fi den
ce in
terv
als
are
used
for
all b
road
and
nar
row
abi
lity /
proc
essin
g cl
uste
rs, t
he
effe
cts
of m
easu
rem
ent e
rror
are
red
uced
furt
her.
a Inf
orm
atio
n ob
tain
ed, i
n pa
rt, f
rom
Wils
on (1
992)
.
Rat
iona
leFo
unda
tion
sA
pplic
atio
ns
Gui
ding
Pri
ncip
les
Ste
p-by
-Ste
pP
roce
ssC
LD
Pop
ulat
ions
Pra
ctic
eR
esea
rch
Test
Dev
elop
men
t
Bri
dge
the
theo
ry-p
ract
ice
gap
Pro
vide
stan
dard
nom
encl
atur
e
Iden
tifi
cati
on o
fco
gnit
ive
proc
essi
ngst
reng
ths
and
wea
knes
s
Impr
ove
the
valid
ity
ofin
telli
genc
e te
sts
Impr
ove
unde
rsta
ndin
g of
re
lati
ons
betw
een
cogn
itiv
e an
d ac
adem
icco
nstr
ucts
Blu
epri
nt fo
rim
prov
ing
upon
the
subs
tant
ive
and
stru
ctur
alva
lidit
y of
tes
ts
Pill
ar 1
CH
C T
heor
y
Pill
ar 2
Bro
ad(S
trat
um I
I)T
est
Cla
ssif
icat
ion
Pill
ar 3
Nar
row
(Str
atum
I)
Tes
tC
lass
ific
atio
ns
Sele
ct b
atte
ryth
at b
est
addr
esse
sre
ferr
al c
once
rns
Sele
ctin
telli
genc
eba
tter
y
Rev
iew
C-L
TC
and
sele
ct t
ests
that
are
like
lyto
be
mos
t fa
ir
Use
clu
ster
sba
sed
on a
ctua
lno
rms
whe
n po
ssib
le
Iden
tify
Bro
adan
d N
arro
w C
HC
abili
ties
mea
sure
dby
bat
tery
Incl
ude
test
s fr
omC
-LT
C n
eede
d fo
rre
ferr
al d
espi
teC
HC
cla
ssif
icat
ion
Sele
ct t
ests
clas
sifi
ed t
hrou
ghan
acc
epta
ble
met
hod
Sele
ct t
ests
to m
easu
reC
HC
abi
litie
sno
t m
easu
red
by b
atte
ry
Adm
inis
ter
enti
reco
llect
ion
of t
ests
sele
cted
inst
anda
rdiz
ed w
ay
Whe
n br
oad
abili
ty is
unde
rrep
rese
nted
,ob
tain
fro
man
othe
r ba
tter
y
Adm
inis
ter
batt
ery
and
supp
lem
enta
l tes
tsas
nec
essa
ry
Use
C-L
IM t
oco
mpa
re r
esul
ts t
oex
pect
ed p
atte
rnof
per
form
ance
Whe
n cr
ossi
ngba
tter
ies
use
test
s de
velo
ped
and
norm
edw
ithi
n a
few
yea
rs
Ent
er d
ata
into
XB
A D
MIA
If p
atte
rn e
vide
ntre
sult
s ar
e in
valid
,ca
nnot
inte
rpre
tda
ta f
urth
er
Sele
ct t
ests
fro
mth
e sm
alle
stnu
mbe
r of
batt
erie
s to
min
imiz
e er
ror
Fol
low
XB
Ain
terp
reti
vegu
idel
ines
If n
o pa
tter
n,re
sult
s ar
e va
lid,
inte
rpre
t vi
aX
BA
gui
delin
es
Fig
ure
1.1
Ove
rvie
w o
f th
e X
BA
Ap
pro
ach
Not
e: C
HC
= C
atte
ll-H
orn-
Car
roll;
C-L
TC =
Cul
ture
-Lan
guag
e Te
st C
lass
ifi ca
tions
; C-L
IM =
Cul
ture
-Lan
guag
e In
terp
retiv
e M
atri
x; C
LD =
C
ultu
rally
and
Lin
guis
tical
ly D
iver
se; X
BA D
MIA
= C
ross
-Bat
tery
Ass
essm
ent D
ata
Man
agem
ent a
nd In
terp
retiv
e A
ssis
tant
.
OVERVIEW 7
three pillars. Together, these pillars (described later in this chapter) pro-vide the knowledge base necessary to organize theory- driven, compre-hensive, reliable, and valid assess-ment of cognitive abilities / pro-cesses.
RATIONALE FOR THE XBA APPROACH
The XBA approach has signifi cant implications for practice, research, and test development. A brief discussion of these implications follows.
Practice
The XBA approach provides “a much needed and updated bridge between current intellectual theory and research and practice” (Flanagan & McGrew, 1997, p. 322). The results of several joint factor analyses conducted over the past 10+ years demonstrated that none of our intelligence batteries contained measures that suffi ciently approximated the full range of broad abilities / pro-cesses that defi ne the structure of intelligence specifi ed in contemporary psychometric theory (e.g., Carroll, 1993; Flanagan & McGrew, 1998; Horn, 1991; Keith, Kranzler, & Flanagan, 2001; McGrew, 1997; Phelps, McGrew, Knopik, & Ford, 2005; Woodcock, 1990). Indeed, the joint factor analyses conducted by Woodcock suggested that it may be necessary to cross batter-ies to measure a broader range of cognitive abilities than that provided by a single intelligence battery.
A summary of the fi ndings of the joint factor analytic studies of intelligence batteries that were published before 2000 are presented in Rapid Reference 1.2. As may be seen in this table, most batteries fell far short of measuring all seven of the broad cognitive abilities / processes listed. Of the major intel-ligence batteries in use prior to 2000, most failed to measure three or more broad CHC abilities (viz., Ga, Glr, Gf, and Gs) that were (and are) considered important in understanding and predicting school achievement. In fact, Gf, often considered to be the essence of intelligence, was either not measured or
DON’T FORGET
The XBA approach allows prac-titioners to reliably measure a wider range (or a more in- depth but selective range) of cognitive abilities / processes than that rep-resented by a single intelligence battery.
Rep
rese
ntat
ion
of B
road
CH
C A
bilit
ies /
Pro
cess
es o
n N
ine
Inte
llig
ence
Bat
teri
es P
ubli
shed
Pri
or to
200
0
Gf
Gc
Gv
Gsm
Glr
Ga
Gs
WIS
C- I
II—
Voca
bula
ry
Info
rmat
ion
Sim
ilari
ties
Com
preh
ensi
on
Bloc
k D
esig
n
Obj
ect A
ssem
bly
Pict
ure
Arr
ange
men
t
Pict
ure
Com
plet
ion
Maz
es
Dig
it Sp
an—
—Sy
mbo
l Sea
rch
Cod
ing
WA
IS- R
—Vo
cabu
lary
Info
rmat
ion
Sim
ilari
ties
Com
preh
ensi
on
Bloc
k D
esig
n
Obj
ect A
ssem
bly
Pict
ure
Com
plet
ion
Pict
ure
Arr
ange
men
t
Dig
it Sp
an—
—D
igit-
Sym
bol
WPP
SI- R
—Vo
cabu
lary
Info
rmat
ion
Sim
ilari
ties
Com
preh
ensi
on
Bloc
k D
esig
n
Obj
ect A
ssem
bly
Pict
ure
Com
plet
ion
Maz
es
Geo
met
ric
Des
ign
Sent
ence
s—
—A
nim
al P
egs
KA
ITM
yste
ry C
odes
Logi
cal S
teps
Defi
niti
ons
Fam
ous
Face
s
Aud
itory
C
ompr
ehen
sion
Dou
ble
Mea
ning
s
Mem
ory
for
Bloc
k D
esig
ns—
Reb
us L
earn
ing
Reb
us D
elay
ed
Rec
all
Aud
itory
Del
ayed
R
ecal
l
——
Rapid
Refer
ence
1.2
K- A
BCM
atri
x A
nalo
gies
—Tr
iang
les
Face
Rec
ogni
tion
Ges
talt
Clo
sure
Mag
ic W
indo
w
Han
d M
ovem
ents
Spat
ial M
emor
y
Phot
o Se
ries
Num
ber
Rec
all
Wor
d O
rder
——
—
CA
S—
—Fi
gure
Mem
ory
Verb
al S
patia
l R
elat
ions
Non
verb
al M
atri
ces
Wor
d Se
ries
Sent
ence
Rep
etiti
on
Sent
ence
Que
stio
ns
——
Mat
chin
g N
umbe
rs
Rec
eptiv
e A
tten
tion
Plan
ned
Cod
es
Num
ber
Det
ectio
n
Plan
ned
Con
nect
ions
Expr
essi
ve
Att
entio
n
DA
SM
atri
ces
Pict
ure
Sim
ilari
ties
Sequ
entia
l and
Q
uant
itativ
e R
easo
ning
Sim
ilari
ties
Verb
al
Com
preh
ensi
on
Wor
d D
efi n
ition
s
Nam
ing
Voca
bula
ry
Patt
ern
Con
stru
ctio
n
Bloc
k Bu
ildin
g
Cop
ying
Mat
chin
g Le
tter
- Li
ke F
orm
s
Rec
all o
f Des
igns
Rec
ogni
tion
of
Pict
ures
Rec
all o
f Dig
itsR
ecal
l of O
bjec
ts—
Spee
d of
In
form
atio
n Pr
oces
sing
(con
tinue
d)
Gf
Gc
Gv
Gsm
Glr
Ga
Gs
WJ-
RC
once
pt F
orm
atio
n
Ana
lysi
s- Sy
nthe
sis
Ora
l Voc
abul
ary
Pict
ure
Voca
bula
ry
List
enin
g C
ompr
ehen
sion
Verb
al A
nalo
gies
Spat
ial R
elat
ions
Pict
ure
Rec
ogni
tion
Vis
ual C
losu
re
Mem
ory
for
Wor
ds
Mem
ory
for
Sent
ence
s
Num
bers
Rev
erse
d
Mem
ory
for
Nam
es
Vis
ual- A
udito
ry
Lear
ning
Del
ayed
Rec
all:
Mem
ory
for
Nam
es
Del
ayed
Rec
all:
Vis
ual- A
udito
ry
Lear
ning
Inco
mpl
ete
Wor
ds
Soun
d Bl
endi
ng
Soun
d Pa
tter
ns
Vis
ual M
atch
ing
Cro
ss O
ut
SB:F
EM
atri
ces
Equa
tion
Build
ing
Num
ber
Seri
es
Verb
al R
elat
ions
Com
preh
ensi
on
Abs
urdi
ties
Voca
bula
ry
Patt
ern
Ana
lysi
s
Bead
Mem
ory
Cop
ying
Mem
ory
for
Obj
ects
Pape
r Fo
ldin
g &
C
uttin
g
Mem
ory
for
Sent
ence
s
Mem
ory
for
Dig
its
——
—
Not
e: C
lass
ifi ca
tions
of t
ests
in t
his
tabl
e ar
e ba
sed
on a
sum
mar
y of
fact
or a
naly
tic r
esea
rch
cond
ucte
d by
Car
roll
(199
3); F
lana
gan
and
McG
rew
(199
8); H
orn
(199
1);
Keith
, Kra
nzle
r, an
d Fl
anag
an (
2001
); M
cGre
w (1
997)
; and
Woo
dcoc
k (1
990)
. WIS
C- I
II =
Wec
hsle
r In
telli
genc
e Sc
ale
for
Chi
ldre
n–T
hird
Edi
tion
(Wec
hsle
r, 19
91);
WA
IS-
R =
Wec
hsle
r A
dult
Inte
llige
nce
Scal
e–R
evis
ed (
Wec
hsle
r, 19
81);
WPP
SI- R
= W
echs
ler
Pres
choo
l and
Pri
mar
y Sc
ale
of In
telli
genc
e–R
evis
ed (
Wec
hsle
r, 19
89);
KA
IT =
Ka
ufm
an A
dole
scen
t and
Adu
lt In
telli
genc
e Te
st (
Kauf
man
& K
aufm
an, 1
993)
; K- A
BC =
Kau
fman
Ass
essm
ent B
atte
ry fo
r C
hild
ren
(Kau
fman
& K
aufm
an, 1
983)
; CA
S =
C
ogni
tive
Ass
essm
ent S
yste
m (
Das
& N
aglie
ri, 1
997)
; DA
S =
Diff
eren
tial A
bilit
y Sc
ales
(El
liott
, 199
0); W
J- R
= W
oodc
ock-
John
son
Psyc
ho- E
duca
tiona
l Bat
tery
–Rev
ised
(W
oodc
ock
& Jo
hnso
n, 1
989)
; SB
:FE
= S
tanf
ord-
Bine
t Int
ellig
ence
Sca
le–F
ourt
h Ed
ition
(T
horn
dike
, Hag
en, &
Sat
tler,
1986
).
Rep
rodu
ced
with
per
mis
sion
from
Gui
lford
.
OVERVIEW 11
not measured adequately by most of the intelligence batteries included in Rapid Reference 1.2 (i.e., WISC- III, WAIS- R, WPPSI- R, K- ABC, and CAS; Alfonso, Flanagan, & Radwan, 2005).
The fi nding that the abilities not
measured by the intelligence batter-ies listed in Rapid Reference 1.2 are important in understanding children’s
learning diffi culties provided the impetus for developing the XBA approach (Flanagan & McGrew 1997). In effect, the XBA approach was developed to systematically replace the dashes in Rapid Reference 1.2 with tests from another battery. As such, this approach guides practitioners in the selection of tests, both core and supplemental, that together provide measurement of abilities / processes that is considered suffi cient in both breadth and depth for the purpose of addressing referral concerns.
Another benefi t of the XBA approach is that it facilitates communication among professionals. Most scientifi c disciplines have a standard nomencla-ture (i.e., a common set of terms and defi nitions) that facilitates communica-tion and guards against misinterpretation. For example, the standard nomen-clature in chemistry is refl ected in the Periodic Table; in biology, it is refl ected in the classifi cation of animals according to phyla; in psychology and psychiatry, it is refl ected in the Diagnostic and Statistical Manual of Mental Disorders; and in medicine, it is refl ected in the International Classifi cation of Diseases. Underlying the XBA approach is a standard nomenclature, or Table of Human Cognitive
Abilities, that includes classifi cations of over 500 tests according to the broad and narrow CHC abilities / processes they measure (see also Alfonso et al., 2005; Flanagan & Ortiz, 2001; Flanagan, McGrew, & Ortiz, 2000; Flanagan, Ortiz, Alfonso, & Mascolo, 2002, 2006). The XBA classifi cation system has had a positive impact on communication among practitioners, has improved research on the relations between cognitive and academic constructs, and has resulted in substantial improvements in the measurement of cognitive constructs, as may be seen in the design and structure of current intelligence batteries (e.g., WJ III, KABC- II, DAS- II, SB5).
Finally, the XBA approach offers practitioners a psychometrically de-
DON’T FORGET
The XBA approach guides prac-titioners in the selection of tests, both core and supplemental, that together provide measurement of abilities / processes that is consid-ered suffi cient in both breadth and depth for the purpose of addressing referral concerns.
12 ESSENTIALS OF CROSS-BATTERY ASSESSMENT
fensible means to identifying population- relative (or normative) strengths and weaknesses in cogni-tive abilities / processes. According to Brackett and McPherson (1996), “the limited capacity of standard-ized instruments to assess isolated cognitive processes creates a ma-jor weakness in intracognitive dis-
crepancy models. Although analysis of [Wechsler] subtests typically report measures of distinct cognitive abilities, such abilities may not emerge by individual subtests but rather in combination with other subtests” (p. 79). The XBA approach addresses this limitation. By focusing interpretations on cognitive ability clusters (i.e., via combinations of construct- relevant sub-tests) that contain qualitatively different indicators of each broad CHC cog-nitive ability / process, the identifi cation of normative processing strengths and weaknesses via XBA procedures is both psychometrically defensible and theoretically sound. In sum, the XBA approach addresses the longstanding need within the entire fi eld of assessment, from learning disabilities to neuro-psychological assessment, for methods that “provide a greater range of infor-mation about the ways individuals learn—the ways individuals receive, store, integrate, and express information” (Brackett & McPherson, p. 80). Because current intelligence tests provide a broader range of information than their predecessors, it is not surprising that results of recent studies demonstrated that specifi c cognitive abilities / processes explain signifi cant variance in aca-demic outcomes (e.g., reading achievement) above and beyond the variance accounted for by g (e.g., Floyd, Keith, Taub, & McGrew, 2006; Vanderwood, McGrew, Flanagan, & Keith, 2002).
Research
The XBA approach was also developed to promote a greater understand-ing of the relationship between cognitive abilities and important outcome criteria. Because XBAs are based on the empirically supported CHC theory and constructed in a psychometrically defensible manner, they represent a valid means of measuring cognitive constructs (Flanagan, 2000; Phelps et al.,
DON’T FORGET
The XBA approach offers practitio-ners a psychometrically defensible means to identifying population- relative (or normative) strengths and weaknesses in cognitive abili-ties / processes.
OVERVIEW 13
2005). It is noteworthy that when second- order constructs are composed of (moderately) correlated but qualitatively distinct measures, they will tend to have higher correlations with complex criteria (e.g., academic achievement), as compared to lower- order constructs, because they are broader in what they measure (Comrey, 1988). Predictive statements about different achieve-ments (i.e., criterion- related inferences) that are made from XBA clusters are based on a more solid foundation than individual subtests (and perhaps some global scores from single intelligence batteries) because the predictor con-structs are represented by relatively pure and qualitatively distinct measures of broad CHC abilities / processes. Thus, improving the validity of CHC abil-ity measures (i.e., intelligence batteries) has further elucidated the relations between CHC cognitive abilities / processes and different achievement and vocational / occupational outcomes (e.g., Flanagan, 2000; Floyd, Bergeron, & Alfonso, 2006; Floyd, Keith, Taub, & McGrew, in press; McGrew, 1997; Vanderwood, McGrew, Flanagan, & Keith, 2002).
Test Development
Although there was substantial evidence of at least eight or nine broad cog-nitive CHC abilities / processes by the late 1980s, the tests of the time did not refl ect this diversity in measurement. For example, Rapid Reference 1.2 shows that the WISC-III, WPPSI- R, K- ABC, KAIT, WAIS- R, and CAS bat-teries only measured two or three broad CHC abilities / processes adequately. The Wechslers primarily measured Gv and Gc. The K- ABC primarily mea-sured Gv and Gsm, and to a much lesser extent Gf, while the KAIT primarily measured Gc, Gf, and Glr, and to a much lesser extent Gv. The CAS measured Gs, Gsm, and Gv. Finally, while the DAS and SB:FE did not provide suffi -cient coverage of abilities to narrow the gap between contemporary theory and practice, their comprehensive measurement of approximately four CHC abilities was nonetheless an improvement over the previously mentioned bat-teries. Rapid Reference 1.2 shows that only the WJ- R included measures of all broad cognitive abilities as compared to the other batteries available at that time. Nevertheless, most of the broad abilities were not measured adequately by the WJ- R (Alfonso et al., 2005; McGrew & Flanagan, 1998).
In general, Rapid Reference 1.2 shows that Gf, Gsm, Glr, Ga, and Gs were not measured well by the majority of intelligence batteries published prior
14 ESSENTIALS OF CROSS-BATTERY ASSESSMENT
to 2000. Therefore, it is clear that most test authors did not use contempo-rary psychometric theories of the structure of cognitive abilities to guide the development of their intelligence batteries. As such, a substantial theory- practice gap existed—that is, theories of the structure of cognitive abilities were far in advance of commonly used intelligence batteries. In fact, prior to the mid- 1980s, theory seldom played a role in intelligence test development. The numerous dashes in Rapid Reference 1.2 exemplify the “theory- practice gap” that existed in the fi eld of intellectual assessment at that time (Alfonso et al., 2005).
In the past decade, Gf- Gc theory, and more recently CHC theory, has had a signifi cant impact on the revision of old and the development of new in-telligence batteries. For example, a wider range of broad and narrow abili-ties / processes is represented on current intelligence batteries than that which was represented on previous editions of these tests. Rapid Reference 1.3 pro-vides several salient examples of the impact that CHC theory and XBA CHC test classifi cations have had on intelligence test development in recent years. This rapid reference lists the major intelligence tests in the order in which they were revised, beginning with those tests with the greatest number of years between revisions (i.e., K- ABC) and ending with newly developed tests (i.e., RIAS and WRIT) and tests that have yet to be revised (e.g., CAS). As is obvious from a review of Rapid Reference 1.3, CHC theory and XBA CHC test classifi cations have had a signifi cant impact on recent test development (Alfonso et al., 2005).
Of the seven intelligence batteries (including both comprehensive and brief measures) that were published since 2000, the test authors of four clearly used CHC theory and XBA CHC test classifi cations as a blueprint for test de-velopment (i.e., WJ III, SB5, KABC- II, and DAS- II), and the test authors of two were obviously infl uenced by CHC theory (i.e., RIAS and WRIT). Only the authors of the Wechsler Scales (i.e., WPPSI- III, WISC- IV, WAIS- III) and CAS did not state explicitly that CHC theory was used as a guide for revi-sion.1 Nevertheless, the authors of the Wechsler Scales acknowledged the re-search of Cattell, Horn, and Carroll in their most recent manuals (Wechsler,
1 Das and Naglieri developed the CAS from PASS theory; therefore, their test is based on an information- processing theory, rather than any specifi c theory within the psychometric tradition.
Impa
ct o
f CH
C T
heor
y an
d X
BA
CH
C T
est C
lass
ifi c
atio
ns o
n In
tell
igen
ce T
est D
evel
opm
ent
Tes
t (Y
ear
of P
ub
licat
ion
)R
evis
ion
(Yea
r o
f Pu
blic
atio
n)
K- A
BC
(19
83)
No
obvi
ous
impa
ct.
KA
BC
- II (
2004
)Pr
ovid
ed a
sec
ond
glob
al s
core
tha
t inc
lude
d fl u
id a
nd c
ryst
alliz
ed a
bilit
ies;
in
clud
ed s
ever
al n
ew s
ubte
sts
mea
suri
ng r
easo
ning
; int
erpr
etat
ion
of te
st
perf
orm
ance
may
be
base
d on
CH
C t
heor
y or
Lur
ia’s
theo
ry; p
rovi
ded
asse
ssm
ent o
f fi v
e C
HC
bro
ad a
bilit
ies /
proc
esse
s. T
hree
add
ition
al C
HC
br
oad
abili
ties /
proc
esse
s ar
e m
easu
red
by it
s co
unte
rpar
t, th
e K
TEA
- II,
repr
esen
ting
8 br
oad
CH
C a
bilit
ies /
proc
esse
s ac
ross
the
se c
o- n
orm
ed
inst
rum
ents
.
SB
:FE
(19
86)
Use
d a
thre
e- le
vel h
iera
rchi
cal m
odel
of t
he s
truc
ture
of c
ogni
tive
abili
ties
to
guid
e co
nstr
uctio
n of
the
test
: the
top
leve
l inc
lude
d a
gene
ral r
easo
ning
fac-
tor,
or g
; the
mid
dle
leve
l inc
lude
d th
ree
broa
d fa
ctor
s ca
lled
crys
talli
zed
abil-
ities
, fl u
id- a
naly
tic a
bilit
ies,
and
shor
t- te
rm m
emor
y; th
e th
ird le
vel i
nclu
ded
mor
e sp
ecifi
c fa
ctor
s in
clud
ing
verb
al r
easo
ning
, qua
ntita
tive
reas
onin
g, a
nd
abst
ract
/ visu
al r
easo
ning
.
SB
5 (2
003)
Use
d C
HC
the
ory
to g
uide
test
dev
elop
men
t; in
crea
sed
the
num
ber
of
broa
d fa
ctor
s fr
om 4
to 6
; inc
lude
d a
Wor
king
Mem
ory
Fact
or b
ased
on
rese
arch
indi
catin
g its
impo
rtan
ce fo
r ac
adem
ic s
ucce
ss.
WA
IS- R
(19
81)
No
obvi
ous
impa
ct.
WA
IS- I
II (
1997
)En
hanc
ed t
he m
easu
rem
ent o
f fl u
id r
easo
ning
by
addi
ng t
he M
atri
x R
ea-
soni
ng s
ubte
st; i
nclu
ded
four
inde
xes
that
mea
sure
spe
cifi c
abi
litie
s / pr
o-
cess
es m
ore
pure
ly t
han
the
trad
ition
al IQ
s pr
ovid
ed in
the
var
ious
W
echs
ler
Scal
es; i
nclu
ded
a W
orki
ng M
emor
y In
dex
base
d on
rec
ent
rese
arch
indi
catin
g its
impo
rtan
ce fo
r ac
adem
ic s
ucce
ss.
Rapid
Refer
ence
1.3
(con
tinue
d)
Tes
t (Y
ear
of P
ub
licat
ion
)R
evis
ion
(Yea
r o
f Pu
blic
atio
n)
WP
PS
I- R
(19
89)
No
obvi
ous
impa
ct.
WP
PS
I- II
I (20
02)
Inco
rpor
ated
mea
sure
s of
Pro
cess
ing
Spee
d th
at y
ield
ed a
Pro
cess
ing
Spee
d Q
uotie
nt b
ased
on
rece
nt r
esea
rch
indi
catin
g th
e im
port
ance
of
Proc
essi
ng S
peed
for
earl
y ac
adem
ic s
ucce
ss; e
nhan
ced
the
mea
sure
men
t of
fl ui
d re
ason
ing
by a
ddin
g th
e M
atri
x R
easo
ning
and
Pic
ture
Con
cept
s su
btes
ts.
WJ-
R (
1989
)U
sed
mod
ern
Gf-
Gc
theo
ry a
s th
e co
gniti
ve m
odel
for
test
dev
elop
men
t; in
clud
ed t
wo
mea
sure
s of
eac
h of
eig
ht b
road
abi
litie
s / pr
oces
ses.
WJ I
II (
2001
)U
sed
CH
C t
heor
y an
d X
BA C
HC
test
cla
ssifi
catio
ns a
s a
“blu
epri
nt”
for
test
dev
elop
men
t; in
clud
ed t
wo
or t
hree
qua
litat
ivel
y di
ffere
nt n
arro
w
abili
ties
for
each
bro
ad a
bilit
y; t
he c
ombi
ned
cogn
itive
and
ach
ieve
men
t ba
tter
ies
of t
he W
J III
incl
ude
9 of
the
10
broa
d ab
ilitie
s su
bsum
ed in
CH
C
theo
ry.
WIS
C- I
II (
1991
)N
o ob
viou
s im
pact
.
WIS
C- I
V (
2003
)El
imin
ated
Ver
bal a
nd P
erfo
rman
ce IQ
s; r
epla
ced
the
Free
dom
from
Dis-
trac
tibili
ty In
dex
with
the
Wor
king
Mem
ory
Inde
x; r
epla
ced
the
Perc
eptu
al
Org
aniz
atio
n In
dex
with
the
Per
cept
ual R
easo
ning
Inde
x; e
nhan
ced
the
mea
sure
men
t of fl
uid
rea
soni
ng b
y ad
ding
the
Mat
rix
Rea
soni
ng a
nd P
ic-tu
re C
once
pts
subt
ests
; enh
ance
d th
e m
easu
rem
ent o
f Pro
cess
ing
Spee
d w
ith t
he a
dditi
on o
f the
Can
cella
tion
subt
est.
DA
S (
1990
)N
o ob
viou
s im
pact
.
DA
S- I
I (20
06)
CH
C t
heor
y w
as u
sed
as a
gui
de fo
r th
e re
visi
on o
f thi
s ba
tter
y. A
s a
resu
lt,
the
DA
S- II
mea
sure
s as
pect
s of
sev
en b
road
cog
nitiv
e ab
ilitie
s / pr
oces
ses.
RIA
S (
2003
)In
clud
ed in
dica
tors
of fl
uid
and
cry
stal
lized
abi
litie
s.
WR
IT (
2002
)D
evel
oped
to b
e co
nsis
tent
with
cur
rent
the
orie
s of
inte
llige
nce;
pro
vide
d Fl
uid
and
Cry
stal
lized
IQs
base
d on
the
Cat
tell-
Hor
n G
f- G
c th
eory
.
CA
S (
1997
)N
o ob
viou
s im
pact
bas
ed o
n PA
SS t
heor
y.
KA
IT (
1993
)In
clud
ed s
ubte
sts
orga
nize
d ac
cord
ing
to t
he w
ork
of H
orn
and
Cat
tell;
pr
ovid
ed F
luid
and
Cry
stal
lized
IQs.
Not
e: K
- ABC
= K
aufm
an A
sses
smen
t Bat
tery
for
Chi
ldre
n (K
aufm
an &
Kau
fman
, 198
3); K
ABC
- II =
Kau
fman
Ass
essm
ent B
atte
ry fo
r C
hild
ren–
Seco
nd E
ditio
n (K
aufm
an
& K
aufm
an, 2
004)
; SB
:FE
= S
tanf
ord-
Bine
t Int
ellig
ence
Sca
le–F
ourt
h Ed
ition
(T
horn
dike
, Hag
en, &
Sat
tler,
1986
); SB
5 =
Sta
nfor
d- Bi
net I
ntel
ligen
ce S
cale
s–Fi
fth
Editi
on
(Roi
d, 2
003)
; WA
IS- R
= W
echs
ler
Adu
lt In
telli
genc
e Sc
ale–
Rev
ised
(W
echs
ler,
1981
); W
AIS
- III
= W
echs
ler
Adu
lt In
telli
genc
e Sc
ale–
Thi
rd E
ditio
n (W
echs
ler,
1997
); W
PPSI
- R =
Wec
hsle
r Pr
esch
ool a
nd P
rim
ary
Scal
e of
Inte
llige
nce–
Rev
ised
(W
echs
ler,
1989
); W
PPSI
- III
= W
echs
ler
Pres
choo
l and
Pri
mar
y Sc
ale
of In
telli
genc
e–T
hird
Edi
-tio
n (W
echs
ler,
2002
); W
J- R
= W
oodc
ock-
John
son
Psyc
ho- E
duca
tiona
l Bat
tery
–Rev
ised
(W
oodc
ock
& Jo
hnso
n, 1
989)
; WJ I
II =
Woo
dcoc
k- Jo
hnso
n III
Tes
ts o
f Cog
nitiv
e A
bilit
ies
(Woo
dcoc
k, M
cGre
w, &
Mat
her,
2001
); W
ISC
- III
= W
echs
ler
Inte
llige
nce
Scal
e fo
r C
hild
ren–
Thi
rd E
ditio
n (W
echs
ler,
1991
); W
ISC
- IV
= W
echs
ler
Inte
llige
nce
Scal
e fo
r C
hild
ren–
Four
th E
ditio
n (W
echs
ler,
2003
); R
IAS
= R
eyno
lds
Inte
llect
ual A
sses
smen
t Sca
les
(Rey
nold
s &
Kam
phau
s, 2
003)
; WR
IT =
Wid
e R
ange
Inte
llige
nce
Test
(G
lutt
ing,
Ada
ms,
& S
hesl
ow, 2
002)
; CA
S =
Cog
nitiv
e A
sses
smen
t Sys
tem
(D
as &
Nag
lieri
, 199
7); K
AIT
= K
aufm
an A
dole
scen
t and
Adu
lt In
telli
genc
e Te
st (
Kauf
man
&
Kauf
man
, 199
3); D
AS
= D
iffer
entia
l Abi
lity
Scal
es (
Ellio
tt, 1
990)
; DA
S- II
= D
iffer
entia
l Abi
lity
Scal
es–S
econ
d Ed
ition
(El
liott
, 200
7).
Rep
rodu
ced
with
per
mis
sion
from
Gui
lford
.
18 ESSENTIALS OF CROSS-BATTERY ASSESSMENT
2002, 2003). Presently, as Rapid Reference 1.3 shows, nearly all comprehen-sive, individually administered intelligence batteries that are used with some regularity subscribe either explicitly or implicitly to CHC theory (Alfonso et al., 2005; Flanagan et al., 2006).
Convergence toward the incorporation of CHC theory is also seen clearly in Rapid Reference 1.4. This table is identical to Rapid Reference 1.2 except it includes all intelligence batteries that were published after 2000, including recent revisions of many of the tests from Rapid Reference 1.2. A comparison of Rapid Reference 1.2 and Rapid Reference 1.4 shows that many of the gaps in measurement of broad cognitive abilities have been fi lled. Specifi cally, the majority of tests published after 2000 now measure four or fi ve broad cogni-tive abilities adequately (see Rapid Reference 1.4) as compared to two or three (see Rapid Reference 1.2). For example, Rapid Reference 1.4 shows that the WISC- IV, WAIS- III, WPPSI- III, KABC- II, SB5, and DAS- II measure four or fi ve CHC broad abilities. The WISC- IV measures Gf, Gc, Gv, Gsm, and Gs while the KABC- II measures Gf, Gc, Gv, and Glr adequately, and to a lesser extent Gsm. The WAIS- III measures Gc, Gv, Gsm, and Gs adequately, and to a lesser extent Gf, while the WPPSI- III measures Gf, Gc, Gv, and Gs adequately. Finally, the SB5 measures four CHC broad abilities adequately (i.e., Gf, Gc, Gv,
Gsm; Alfonso et al., 2005) and the DAS- II measures fi ve CHC broad abilities adequately (i.e., Gf, Gc, Gv, Gsm, and Glr) and to a lesser extent, Ga and Gs.
Rapid Reference 1.4 shows that the WJ III continues to include measures of all the major broad cognitive abilities / processes and now measures them well, particularly when it is used in conjunction with the Diagnostic Supplement (DS; Woodcock, McGrew, Mather, & Schrank, 2003). Third, a comparison of Rapid References 1.2 and 1.4 indicates that two broad abilities / processes not measured by many intelligence batteries prior to 2000 are now measured by the majority of intelligence batteries available today; that is, Gf and Gsm. These broad abilities / processes may be better represented on revised and new intelligence batteries because of the accumulating research evidence re-garding their importance in overall academic success (see Chapter 2). Finally, Rapid Reference 1.4 reveals that intelligence batteries continue to fall short in their measurement of three CHC broad abilities / processes; specifi cally, Glr, Ga, and Gs. In addition, current intelligence batteries do not provide adequate measurement of most specifi c or narrow CHC abilities / processes, many of which are important in predicting academic achievement. Thus,
Rep
rese
ntat
ion
of B
road
CH
C A
bilit
ies /
Pro
cess
es o
n N
ine
Inte
llig
ence
Bat
teri
es P
ubli
shed
Aft
er 2
000
Gf
Gc
Gv
Gsm
Glr
Ga
Gs
WIS
C- I
VM
atri
x R
easo
ning
Pict
ure
Con
cept
s
Voca
bula
ry
Info
rmat
ion
Sim
ilari
ties
Com
preh
ensi
on
Wor
d R
easo
ning
Bloc
k D
esig
n
Pict
ure
Com
plet
ion
Dig
it Sp
an
Lett
er- N
umbe
r Se
quen
cing
——
Sym
bol S
earc
h
Cod
ing
Can
cella
tion
WA
IS- I
IIaM
atri
x R
easo
ning
Voca
bula
ry
Info
rmat
ion
Sim
ilari
ties
Com
preh
ensi
on
Bloc
k D
esig
n
Obj
ect A
ssem
bly
Pict
ure
Arr
ange
men
t
Pict
ure
Com
plet
ion
Dig
it Sp
an
Lett
er- N
umbe
r Se
quen
cing
——
Sym
bol S
earc
h
Dig
it- Sy
mbo
l Cod
ing
WPP
SI- I
IIM
atri
x R
easo
ning
Pict
ure
Con
cept
s
Voca
bula
ry
Info
rmat
ion
Sim
ilari
ties
Com
preh
ensi
on
Rec
eptiv
e Vo
cabu
lary
Pict
ure
Nam
ing
Wor
d R
easo
ning
Bloc
k D
esig
n
Obj
ect A
ssem
bly
Pict
ure
Com
plet
ion
——
—C
odin
g
Sym
bol S
earc
h
Rapid
Refer
ence
1.4
(con
tinue
d)
Gf
Gc
Gv
Gsm
Glr
Ga
Gs
KA
BC- I
IPa
tter
n R
easo
ning
Stor
y C
ompl
etio
n
Expr
essi
ve
Voca
bula
ry
Verb
al K
now
ledg
e
Rid
dles
Face
Rec
ogni
tion
Tria
ngle
s
Ges
talt
Clo
sure
Rov
er
Bloc
k C
ount
ing
Con
cept
ual T
hink
ing
Num
ber
Rec
all
Wor
d O
rder
Han
d M
ovem
ents
Atla
ntis
Reb
us
Atla
ntis
Del
ayed
Reb
us D
elay
ed
——
WJ I
IIC
once
pt F
orm
atio
n
Ana
lysi
s- Sy
nthe
sis
Verb
al
Com
preh
ensi
on
Gen
eral
Info
rmat
ion
Spat
ial R
elat
ions
Pict
ure
Rec
ogni
tion
Plan
ning
Mem
ory
for
Wor
ds
Num
bers
Rev
erse
d
Aud
itory
Wor
king
M
emor
y
Vis
ual- A
udito
ry
Lear
ning
Ret
riev
al F
luen
cy
Vis
ual- A
udito
ry
Lear
ning
Del
ayed
Rap
id P
ictu
re
Nam
ing
Soun
d Bl
endi
ng
Aud
itory
Att
entio
n
Inco
mpl
ete
Wor
ds
Vis
ual M
atch
ing
Dec
isio
n Sp
eed
Pair
Can
cella
tion
SB5
Non
verb
al F
luid
R
easo
ning
Verb
al F
luid
R
easo
ning
Non
verb
al
Qua
ntita
tive
Rea
soni
ng
Verb
al Q
uant
itativ
e R
easo
ning
Non
verb
al
Kno
wle
dge
Verb
al K
now
ledg
e
Non
verb
al V
isua
l- Sp
atia
l Pro
cess
ing
Verb
al V
isua
l- Spa
tial
Proc
essi
ng
Non
verb
al W
orki
ng
Mem
ory
Verb
al W
orki
ng
Mem
ory
——
—
DA
S- II
Mat
rice
s
Pict
ure
Sim
ilari
ties
Qua
ntita
tive
Rea
soni
ng
Earl
y N
umbe
r C
once
pts
Nam
ing
Voca
bula
ry
Wor
d D
efi n
ition
s
Verb
al
Com
preh
ensi
on
Verb
al S
imila
ritie
s
Patt
ern
Con
stru
ctio
n
Rec
all o
f Des
igns
Rec
ogni
tion
of
Pict
ures
Cop
ying
Mat
chin
g Le
tter
- Lik
e Fo
rms
Rec
all o
f Dig
its-
Forw
ard
Rec
all o
f Dig
its-
Back
war
d
Rec
all o
f Seq
uent
ial
Ord
er
Verb
al M
emor
y
Rap
id N
amin
g
Rec
all o
f Obj
ects
- Im
med
iate
Rec
all o
f Obj
ects
- D
elay
ed
Spee
d of
In
form
atio
n Pr
oces
sing
—
RIA
S—
Gue
ss W
hat
Verb
al R
easo
ning
Wha
t’s M
issi
ng
Odd
- Ite
m O
ut
Non
verb
al M
emor
y
Verb
al M
emor
y—
——
WR
ITM
atri
ces
Verb
al A
nalo
gies
Voca
bula
ry
Dia
mon
ds—
——
—
Not
e: C
HC
cla
ssifi
catio
ns a
re b
ased
on
the
liter
atur
e an
d pr
imar
y so
urce
s su
ch a
s C
arro
ll (1
993)
, Fla
naga
n an
d O
rtiz
(20
01),
Flan
agan
, Ort
iz, A
lfons
o, a
nd M
asco
lo (
2006
), H
orn
(199
1), K
eith
, Fin
e, T
aub,
Rey
nold
s, a
nd K
ranz
ler
(200
6), M
cGre
w (1
997)
, and
McG
rew
and
Fla
naga
n (1
998)
.W
ISC
- IV
= W
echs
ler
Inte
llige
nce
Scal
e fo
r C
hild
ren–
Four
th E
ditio
n (W
echs
ler,
2003
); W
AIS
- III
= W
echs
ler
Adu
lt In
telli
genc
e Sc
ale–
Thi
rd E
ditio
n (W
echs
ler,
1997
); W
PPSI
- III
= W
echs
ler
Pres
choo
l and
Pri
mar
y Sc
ale
of In
telli
genc
e–T
hird
Edi
tion
(Wec
hsle
r, 20
02);
KA
BC- I
I = K
aufm
an A
sses
smen
t Bat
tery
for
Chi
ldre
n–Se
cond
Edi
tion
(Kau
fman
& K
aufm
an, 2
004)
; WJ I
II =
Woo
dcoc
k- Jo
hnso
n III
Tes
ts o
f Cog
nitiv
e A
bilit
ies
(Woo
dcoc
k, M
cGre
w, &
Mat
her
2001
); SB
5 =
Sta
nfor
d- Bi
net I
ntel
ligen
ce S
cale
s–Fi
fth
Editi
on (
Roi
d, 2
003)
; DA
S- II
= D
iffer
entia
l Abi
lity
Scal
es–S
econ
d Ed
ition
(El
liott
, 200
7); R
IAS
= R
eyno
lds
Inte
llect
ual A
sses
smen
t Sca
les
(Rey
nold
s &
Kam
phau
s, 20
03);
WR
IT =
Wid
e R
ange
Inte
llige
nce
Test
(G
lutt
ing,
Ada
ms,
& S
hesl
ow, 2
002)
.a A
lthou
gh t
he W
AIS
- III
was
pub
lishe
d in
199
7, it
is in
clud
ed in
thi
s ta
ble
beca
use
its p
rede
cess
or, t
he W
echs
ler
Adu
lt In
telli
genc
e Sc
ale–
Rev
ised
, was
incl
uded
in R
apid
Ref
-er
ence
1.2
and
in o
rder
to p
rese
nt a
ll re
vise
d W
echs
ler
Scal
es in
one
tab
le.
Rep
rodu
ced
with
per
mis
sion
from
Gui
lford
.
22 ESSENTIALS OF CROSS-BATTERY ASSESSMENT
although there is greater coverage of CHC broad abilities / processes now than there was just a few years ago, the need for the XBA approach to assessment remains (Alfonso et al., 2005).
THE THREE PILLARS OF THE XBA APPROACH
The three pillars of the XBA ap-proach include contemporary CHC theory and the broad and narrow CHC ability classifi cations of all subtests that comprise current cog-nitive and achievement batteries as well as numerous special purpose tests. Each pillar is defi ned briefl y in the following sections and in Rapid Reference 1.5.
The First Pillar of the XBA Approach: CHC Theory
The CHC theory was selected to guide assessment and interpreta-tion because it is based on a more
thorough network of validity evidence than any other contemporary multi-dimensional model of intelligence within the psychometric tradition (see McGrew, 2005; Messick, 1992; Sternberg & Kaufman, 1998). According to Daniel (1997), the strength of the multiple (CHC) cognitive abilities model is that it was arrived at “by synthesizing hundreds of factor analyses con-ducted over decades by independent researchers using many different col-lections of tests. Never before has a psychometric ability model been so fi rmly grounded in data” (pp. 1042–1043). Because nearly all current intel-ligence batteries are based on CHC theory, it will not be described in detail in
DON’T FORGET
Nearly all comprehensive, indi-vidually administered intelligence batteries that are used with some regularity subscribe either explicitly or implicitly to CHC theory.
Three Pillars of the XBA Approach
• The fi rst pillar of the approach is a relatively complete taxonomic framework for describing the structure and nature of cognitive abilities. This taxonomy is the Cattell- Horn- Carroll theory of cognitive abilities (CHC theory).
• The second pillar of the ap-proach is the CHC broad (stra-tum II) classifi cations of cognitive and achievement tests.
• The third pillar of the approach is the CHC narrow (stratum I) classi-fi cations of cognitive and achieve-ment tests.
Rapid Reference 1.5
OVERVIEW 23
this chapter. For a detailed presentation of CHC theory and comprehensive defi nitions of all broad and narrow CHC abilities / processes, see Appendix A of this book.
The Second Pillar of the XBA Approach: CHC Broad (Stratum II) Classifi cations of Cognitive and Achievement Tests
Based on the results of a series of cross- battery confi rmatory factor analysis studies of the major intelligence batteries and the task analyses of many intel-ligence test experts, Flanagan and colleagues classifi ed all the subtests of the major intelligence and achievement batteries according to the particular CHC broad abilities / processes they measured (e.g., Flanagan et al., 2006). To date, well over 500 CHC broad ability classifi cations have been made based on the results of these studies. These classifi cations of cognitive and achievement tests assist practitioners in identifying measures that assess the various broad and narrow abilities / processes represented in CHC theory. Classifi cation of tests at the broad ability / processing level is necessary to improve upon the validity of cognitive assessment and interpretation. Specifi cally, broad abil-ity classifi cations ensure that the CHC constructs that underlie assessments are minimally affected by construct-irrelevant variance (Messick, 1989, 1995). In other words, knowing what tests measure what abilities / processes enables clinicians to organize tests into construct-relevant clusters—clusters that con-tain only measures that are relevant to the construct of interest.
To clarify, construct- irrelevant variance is present when an “assessment is too broad, containing excess reliable variance associated with other distinct con-structs . . . that affects responses in a manner irrelevant to the interpreted constructs” (Messick, 1995, p. 742). For example, the WAIS- III Verbal IQ (VIQ) has construct- irrelevant variance because, in addition to its four indicators of Gc (i.e., Information, Similarities, Vocabulary, Comprehension), it has one indicator of Gq (i.e., Arithmetic) and one indicator of Gsm (i.e., Digit Span). Therefore, the VIQ is a mixed measure of three distinct, broad CHC abilities / processes (Gc, Gq, and Gsm); it contains reliable variance (associated with Gq and Gsm) that is irrelevant to the construct intended to be interpreted (i.e., Gc; McGrew & Flanagan, 1998). The Wechsler VIQ represents a grouping together of sub-tests on the basis of face validity (e.g., grouping tests together that appear to measure the same common concept), an inappropriate aggregation of subtests
24 ESSENTIALS OF CROSS-BATTERY ASSESSMENT
that can actually decrease reliability and validity (Epstein, 1983). The purest Gc composite on the WAIS- III is the Verbal Comprehension Index, because it contains only construct- relevant variance.
Construct- irrelevant variance can also operate at the subtest (as opposed to composite) level. For example, a Verbal Analogies test (e.g., Sun is to day as moon is to ) measures both Gc and Gf. That is, in theory- driven factor- analytic stud-ies, Verbal Analogies tests have sig-nifi cant loadings on both the Gc and Gf factors (e.g., Woodcock, 1990). Therefore, this test is considered factorially complex—a condition that complicates interpretation (e.g., Is poor performance due to low vo-cabulary knowledge [Gc] or poor reasoning ability [Gf ], or both?).
In short, “[A]ny test that mea-sures more than one common fac-tor to a substantial degree yields scores that are psychologically ambiguous and very diffi cult to interpret” (Guilford, 1954, p. 356; cited in Briggs & Cheek, 1986). In-terpretation is far less complicated when composites are derived from relatively pure measures of the underlying construct. Therefore, XBAs are typically designed using only empirically strong or moderate
DON’T FORGET
Invalidity in Assessment
Construct- irrelevant variance—excess reliable variance associated with other distinct constructs that affects responses in a manner irrel-evant to the interpreted construct.The XBA approach guards against this major source of invalidity in assessment by ensuring that only validated measures of a cognitive construct are included in an XBA designed to measure that con-struct.The XBA DMIA organizes the subtests of the major intelligence batteries according to the broad abilities / processes they measure to assist practitioners in designing as-sessments that measure constructs validly.
C A U T I O N
Clusters that contain construct- irrelevant variance are psycho-logically ambiguous and diffi cult to interpret. For example, the traditional Wechsler VIQ contained variance (Gq) that was irrelevant to the construct intended to be interpreted (Gc). The Verbal Com-prehension Index eliminated the ir-relevant Gq variance and, therefore, represented a purer measure of Gc as compared to the VIQ.
OVERVIEW 25
(but not factorially complex or mixed) measures of CHC abilities / processes, following the information presented in Appendix B.2
The Third Pillar of the XBA Approach: CHC Narrow (Stratum I) Classifi cations of Cognitive and Achievement Tests
Narrow ability / processing classifi cations were originally reported in Mc-Grew (1997), then later reported in McGrew and Flanagan (1998) and Flana-gan et al. (2000) following minor modifi cations. Flanagan and her colleagues continued to gather content validity data on cognitive tests and expanded their analyses recently to include tests of academic achievement (Flanagan et al., 2002, 2006). Classifi cations of cognitive tests according to content, format, and task demand at the narrow (stratum I) ability / processing level were necessary to improve further upon the validity of intellectual assess-ment and interpretation (see Messick, 1989). Specifi cally, these narrow ability classifi cations were necessary to ensure that the CHC constructs that under-lie assessments are well represented. According to Messick (1995), construct underrepresentation is present when an “assessment is too narrow and fails to include important dimensions or facets of the construct” (p. 742).
Interpreting the WJ III Concept Formation (CF) subtest as a measure of Fluid Intelligence (i.e., the broad Gf ability / process) is an example of con-struct underrepresentation. This is because CF measures one narrow aspect of Gf (viz., Inductive Reasoning). At least one other Gf measure (i.e., subtest) that is qualitatively different from Inductive Reasoning is necessary to include in an assessment to ensure adequate representation of the Gf construct (e.g., a measure of General Sequential [or Deductive] Reasoning). Two or more
2 Classifi cations of cognitive ability tests as strong, moderate, or mixed measures of CHC abilities were based on the following criteria: A classifi cation of strong was given to a test that had a substantial factor loading (> .50) on a primary factor and a secondary factor loading (if present) that was equal to or less than ½ of its loading on the primary factor. A classifi cation of moderate was given to a test that had a primary factor loading of < .50 and a secondary factor loading (if present) that was less than ½ of the primary loading, or any primary factor loading with a secondary loading between ½ and 7⁄10 of the primary loading. A classifi cation of mixed was given to a test that had a factor loading on a secondary factor that was greater than 7⁄10 of its loading on the primary factor. These criteria were derived from Woodcock (1990).
26 ESSENTIALS OF CROSS-BATTERY ASSESSMENT
qualitatively different indicators (i.e., measures of two or more nar-row abilities / processes subsumed by the broad ability / process) are needed for appropriate construct representation (see Comrey, 1988; Messick, 1989, 1995). The aggregate of CF (a measure of Inductive Rea-soning at the narrow ability level) and the WJ III Analysis- Synthesis test (a measure of General Sequen-tial Reasoning at the narrow ability level), for example, would provide an adequate estimate of the broad Gf construct because these tests are strong measures of Gf and repre-sent qualitatively different aspects of Gf (see Appendix B).
The Verbal Comprehension In-dex (VCI) of the WAIS- III is an example of good construct repre-
sentation. This is because the VCI includes Vocabulary (VL), Similarities (LD / VL), Comprehension (LD), and Information (K0), all of which rep-resent qualitatively different aspects of Gc. Most intelligence batteries yield construct- relevant composites, although some of these composites underrep-resent the broad ability intended to be measured. This is because construct underrepresentation can also occur when the composite consists of two or more measures of the same narrow (stratum I) ability / process. For example, the Number Recall and Word Order subtests of the KABC- II were intended to be interpreted as a representation of the broad Gsm ability / process (Kaufman & Kaufman, 2004). However, these subtests primarily measure Memory Span, a narrow ability / process subsumed by Gsm. Thus, the Gsm cluster of the KABC- II is most appropriately interpreted as Memory Span (a narrow abil-ity / process) rather than an estimate of the broad Gsm ability / process.
“A scale [or broad CHC ability cluster] will yield far more information—and, hence, be a more valid measure of a construct—if it contains more differ-
DON’T FORGET
Invalidity in Assessment
Construct underrepresentation—present when an assessment is too narrow and fails to include important dimensions or facets of a construct.The XBA approach guards against this major source of invalidity in as-sessment by ensuring that at least two different components of a cog-nitive construct are included in an XBA cluster designed to measure that construct.The XBA DMIA organizes the subtests of the major intelligence batteries according to the narrow abilities / processes they measure to assist practitioners in designing as-sessments that measure construct validly.
OVERVIEW 27
entiated items [or tests]” (Clarke & Watson, 1995, p. 316). Cross- battery assessments circumvent the misinterpretations that can result from under-represented constructs by specifying the use of two or more qualitatively different indicators to represent each broad CHC ability / process. In order to ensure that qualitatively different aspects of broad abilities / processes are represented in assessment, classifi cation of cognitive and achievement tests at the narrow (stratum I) ability / processing level was necessary. The subtests of current intelligence batteries, special purpose tests, and comprehensive achievement batteries are classifi ed at both the broad and narrow ability / pro-cessing levels throughout this book (see Appendix B for a summary).
In sum, the latter two XBA pillars guard against two ubiquitous sources of invalidity in assessment: construct- irrelevant variance and construct underrepresentation. Taken together, the three pillars underlying the XBA approach provide the necessary foundation from which to organize assess-ments of cognitive and achievement constructs that are more theoretically driven, comprehensive, and valid.
Prior to discussing the applications of the XBA approach, it is necessary to highlight the various ways in which the approach has evolved since the fi rst edition of this book. As noted earlier, nearly all frequently used intelligence batteries have been revised in recent years. Additionally, these revisions are among the most substantial of the major intelligence batteries in the his-tory of intellectual assessment. As a result, nearly all intelligence batteries include measurement of a broader range of cognitive constructs and, indeed, constructs from a single psychometric theory—CHC theory. Because intel-ligence batteries are substantially better than their predecessors from both a psychometric and theoretical standpoint, the application of XBA methods is less involved. Specifi cally, the mechanics of the approach are more simplistic and may be carried out effortlessly using the automated program included on the CD- ROM accompanying this book (this program hereafter referred to as “XBA DMIA,” which stands for Cross- Battery Assessment Data Man-agement and Interpretive Assistant). Additionally, the interpretation of test performance is enhanced by the XBA DMIA as well as by the interpretive statements that correspond to this program’s output. That is, we provide inter-pretive statements (that may be used verbatim in a psychoeducational report) for every possible outcome of a broad or narrow XBA cluster calculated by the XBA DMIA (see Chapter 3). Rapid Reference 1.6 lists the major changes
New Features of the XBA Approach
1. More easily incorporates and integrates all current intelligence batteries (i.e., WISC- IV, WAIS- III, WPPSI- III, KABC- II, WJ III, SB5, and DAS- II), numerous spe-cial purpose tests, and tests of academic achievement.
2. Uses core tests (and supplemental tests as may be necessary) from a single battery, rather than selected components of a battery, as part of the assessment because (a) current intelligence tests have better representation of the broad CHC abilities / processes and use only two or three subtests to represent them; and (b) the broad abilities / processes measured by current intelligence batteries are typi-cally represented by qualitatively different indicators that are relevant only to the broad ability / processes intended to be measured.
3. Uses actual norms provided by the test’s publisher for CHC broad ability clusters when available.
4. Places greater emphasis on narrow CHC abilities / processes as supported by re-search linking them to acquisition and development of specifi c academic skills.
5. Includes an automated program called Cross- Battery Assessment Data Manage-ment and Interpretive Assistant (XBA DMIA) (on the CD- ROM that accompanies this book), that incorporates and integrates all features of the XBA approach. For example, the XBA DMIA• Incorporates and integrates components of prevailing interpretive systems of
the major intelligence batteries, including optional clinical clusters unique to WISC- IV, WAIS- III, and SB5.
• Calculates CHC broad and narrow ability / processing clusters that are gener-ated from either two or three individual subtests.
• Graphs data to provide a pictorial representation of all interpretable broad and narrow ability / processing clusters and the subtests that comprise them.
6. Includes interpretive statements for all possible outcomes regarding data from two or three subtest combinations for broad and narrow ability / processing areas.
7. Expands coverage of CHC theory to include abilities typically measured on achievement tests (e.g., Broad Reading and Writing [Grw] , Quantitative Knowl-edge [Gq] , and extended components of Auditory Processing [Ga]), providing additional information useful in the identifi cation of specifi c learning disability (SLD).
8. Incorporates the identifi cation of disorders in basic psychological processes in the interpretive system in a manner consistent with the defi nition of SLD in IDEA 2004 and includes an automated program called SLD Assistant.
9. Includes advancements to the interpretive system of the Culture- Language Inter-pretive Matrix used with culturally and linguistically diverse individuals.
10. Includes an automated program called Culture- Language Interpretive Matrix (C- LIM), which calculates and graphs results to facilitate decision making as it pertains to differentiating difference from disability with individuals from cultur-ally and linguistically diverse backgrounds.
Rapid Reference 1.6
OVERVIEW 29
that have taken place in the XBA approach since the publication of the fi rst edition of this book.
APPLICATION OF THE XBA APPROACH
Guiding Principles
In order to ensure that XBA procedures are psychometrically and theoreti-cally sound, it is recommended that practitioners adhere to several guiding principles. These principles were listed previously in Figure 1.1 and are de-fi ned briefl y in the following section.
First, select an intelligence battery that best addresses referral concerns. It is expected that the battery of choice will be one that is deemed most re-sponsive to referral concerns. These batteries may include, but are certainly not limited to, the Wechsler Scales, WJ III, SB5, KABC- II, and DAS- II. It is important to note that the use of conormed tests, such as the WJ III tests of cognitive ability and tests of achievement and the KABC- II and KTEA- II, may allow for the widest coverage of broad and narrow CHC abilities / pro-cesses.
Second, use subtests and clusters / composites from a single battery whenever possible to represent broad CHC abilities / processes. In other words, best practices involve using actual norms whenever they are available in lieu of arithmetic averages of scaled scores from different batteries. In the past, it was necessary to convert subtest- scaled scores from different batteries to a common metric (using the table in Appendix E, for example) and then average them (after determining that there was a nonsignifi cant difference between the scores) in order to build construct relevant broad CHC abil-ity / processing clusters. Because the development of current intelligence bat-teries benefi ted greatly from current CHC theory and research, this practice is seldom necessary at the broad ability / processing level. It continues to be necessary when testing hypotheses about aberrant performance within broad ability / processing domains and when measurement of narrow abilities / pro-cesses is deemed necessary (see Chapters 2 and 3).
Third, when constructing CHC broad and narrow ability clusters, select tests that have been classifi ed through an acceptable method, such as through CHC theory- driven factor analyses or expert consensus content- validity
30 ESSENTIALS OF CROSS-BATTERY ASSESSMENT
studies. All test classifi cations included in the works of Flanagan and col-leagues have been classifi ed through these acceptable methods (Flanagan & Ortiz, 2001; Flanagan et al., 2006). For example, when constructing broad (stratum II) ability / processing clusters, relatively pure CHC indicators should be included (i.e., tests that had either strong or moderate [but not mixed] load-ings on their respective factors in theory driven factor analyses). Further-more, to ensure appropriate construct representation when constructing broad (stratum II) ability / processing clusters, two or more qualitatively differ-
ent narrow (stratum I) ability / processing indicators should be included to represent each domain. Without empirical classifi cations of tests, constructs may not be adequately represented and, therefore, inferences about an in-dividual’s broad (stratum II) ability / process cannot be made. Of course, the more broadly a construct is represented (i.e., through the derivation of a cluster based on multiple qualitatively different narrow ability / processing indicators), the more confi dence one has in drawing inferences about the ability / process presumed to underlie it. A minimum of two qualitatively dif-ferent indicators per CHC cluster is recommended in the XBA approach for practical reasons (viz., time effi cient assessment).
Fourth, when at least two qualitatively different indicators of a broad abil-ity / process of interest is not available on the core battery, then supplement the core battery with at least two qualitatively different indicators of that broad ability from another battery. In other words, if an evaluator is inter-ested in measuring Auditory Processing (Ga), and the core battery includes either one or no Ga subtests, then select a Ga cluster from another battery to supplement the core battery.
Fifth, when crossing batteries (e.g., augmenting a core battery with rele-vant CHC clusters from another battery) or when constructing CHC broad or narrow ability / processing clusters using tests from different batteries (e.g., averaging scores when the construct of interest is not available on a single bat-tery), select tests that were developed and normed within a few years of one another to minimize the effect of spurious differences between test scores that may be attributable to the “Flynn effect” (Flynn, 1984). The subtests listed in the XBA DMIA are from batteries and tests that were normed within 10 years of one another.
Sixth, select tests from the smallest number of batteries to minimize the effect of spurious differences between test scores that may be attributable to
OVERVIEW 31
differences in the characteristics of independent norm samples (McGrew, 1994). In most cases, using select tests from a single battery to augment the constructs measured by any other major intelligence battery is suffi cient to represent the breadth of broad cognitive abilities / processes adequately as well as to allow for at least three qualitatively different narrow ability / pro-cessing indicators of most CHC cognitive constructs.
Noteworthy is the fact that when the XBA guiding principles are imple-mented systematically and the recommendations for development, use, and interpretation of clusters are adhered to, the potential error introduced through the crossing of norm groups is negligible (Flanagan & Ortiz, 2001; McGrew & Flanagan, 1998). Furthermore, although there are other limita-tions to crossing batteries, this systematic approach to the assessment and interpretation of cognitive abilities / processes has far fewer implications with regard to the potential for error than those associated with the improper use and interpretation of cognitive performance inherent in many currently used assessment approaches (e.g., subtest analysis, discrepancy analysis, atheoreti-cal approaches to assessment and interpretation, and so forth).
IMPLEMENTATION OF THE XBA APPROACH STEP- BY- STEP
The XBA approach may be carried out following a straightforward set of steps. These steps are outlined in Rapid Reference 1.7 and described in further detail in Chapter 2.
USE OF THE XBA APPROACH WITH CULTURALLY AND LINGUISTICALLY DIVERSE POPULATIONS
Application of the XBA approach with diverse individuals rests on
XBA Step- by- Step
1. Select primary intelligence bat-tery for assessment.
3. Select tests to measure CHC abilities / processes not measured by primary battery.
4. Administer primary battery and any supplemental tests as neces-sary.
5. Enter data into the XBA DMIA.6. Follow XBA guidelines pre-
sented in Chapter 3 to interpret XBA DMIA output.
Rapid Reference 1.7
32 ESSENTIALS OF CROSS-BATTERY ASSESSMENT
the premise that an empirically based selection of tests, known to represent particular constructs, coupled with a consideration of the relevant cultural and linguistic dimensions of such tests, can provide more reliable, valid, and interpretable data than that ordinarily obtained using traditional methods. Careful and deliberate selection of tests, based on factors relevant to the back-ground of the individual being assessed, creates a unique battery of tests that is responsive to the particular referral questions. Using the XBA approach, practitioners can develop custom batteries for individuals of culturally and linguistically diverse backgrounds that differ as a function of both the specifi c language competencies and the cultural experiences of the individual, as well as the specifi c nature of the referral concerns. With respect to issues of bias re-lated to test selection, the basic goal in constructing XBAs for use with diverse individuals is to ensure a balance between empirical issues and considerations related to cultural and linguistic factors. The construction of an appropriate XBA for use with diverse individuals is presented in Chapter 5 along with a detailed explanation of how to interpret their test performances.
CONCLUSIONS
Recent refi nements to the XBA approach, including automating the process, have made this method of assessment both practical and easy to implement. Its continued popularity revolves around its use in the identifi cation of students with specifi c learning disability (Chapter 4) and in assisting in the process of determining difference from disability in students from culturally and linguisti-cally diverse backgrounds (Chapter 5). This is because the XBA approach (a) allows for fl exibility in designing assessment batteries to meet the unique needs of the individual; (b) provides a defensible interpretive method for identifying cognitive ability / processing strengths and weaknesses (impor-tant in the evaluation of learning disabilities); and (c) is systematic, specifying steps for evaluating the cognitive capabilities of individuals with learning needs, including those from diverse cultural and linguistic backgrounds.
REFERENCES
Alfonso, V. C., Flanagan, D. P., & Radwan, S. (2005). The impact of the Cattell- Horn-Carroll Theory on test development and interpretation of cognitive and academic
OVERVIEW 33
abilities. In D. P. Flanagan & P. L. Harrison (Eds.), Contemporary intellectual assess-
ment: Theories, tests, and issues (2nd ed., pp. 185–202). New York: Guilford.Brackett, J., & McPherson, A. (1996). Learning disabilities diagnosis in postsecondary
students: A comparison of discrepancy- based diagnostic models. In N. Gregg, C. Hoy, & A. F. Gay (Eds.), Adults with learning disabilities: Theoretical and practical per-
spectives (pp. 68–84). New York: Guilford.Briggs, S. R., & Cheek, J. M. (1986). The role of factor analysis in the development and
evaluation of personality scales [Special Issue: Methodological developments in personality research]. Journal of Personality, 54(1), 106–148.
Carroll, J. B. (1993). Human cognitive abilities: A survey of factor- analytic studies. Cambridge, UK: Cambridge University Press.
Carroll, J. B. (1997). The three- stratum theory of cognitive abilities. In D. P. Flanagan, J. L. Genshaft, & P. L. Harrison (Eds.), Contemporary intellectual assessment: Theories,
tests, and issues (pp. 122–130). New York: Guilford.Carroll, J. B. (1998). Foreword. In K. S. McGrew & D. P. Flanagan, The intelligence test
Clarke, L. A., & Watson, D. (1995). Constructing validity: Basic issues in objective scale development. Psychological Assessment, 7, 309–319.
Comrey, A. L. (1988). Factor- analytic methods of scale development in personality and clinical psychology. Journal of Consulting and Clinical Psycholog y, 56(5), 754–761.
Daniel, M. H. (1997). Intelligence testing: Status and trends. American Psychologist, 52, 1038–1045.
Das, J. P., & Naglieri, J. A. (1997). Das- Naglieri Cognitive Assessment System. Itasca, IL: Riverside Publishing.
Dehn, M. J. (2006). Essentials of processing assessment. New York: Wiley.Elliot, C. (1990). Differential Ability Scales (DAS). San Antonio, TX: The Psychological
Corporation.Elliot, C. (2007). Differential Ability Scales–Second Edition (DAS- II). San Antonio, TX:
PsychCorp.Epstein, S. (1983). Aggression and beyond: Some basic issues on the prediction of be-
havior. Journal of Personality, 51, 360–392.Fiorello, C. A., & Hale, J. B. (2006). Cognitive hypothesis testing and response to in-
terventions for children with reading problems. Psycholog y in the Schools, 43, 835–853.Flanagan, D. P. (2000). Wechsler- based CHC cross- battery assessment and reading
achievement: Strengthening the validity of interpretations drawn from Wechsler test scores. School Psycholog y Quarterly, 15(3), 295–329.
Flanagan, D. P., & McGrew, K. S. (1997). A cross- battery approach to assessing and interpreting cognitive abilities: Narrowing the gap between practice and cognitive science. In D. P. Flanagan, J. L. Genshaft, & P. L. Harrison (Eds.), Contemporary in-
tellectual assessment: Theories, tests, and issues (pp. 314–325). New York: Guilford.Flanagan, D. P., & McGrew, K. S. (1998). Interpreting intelligence tests from contem-
porary Gf- Gc theory: Joint confi rmatory factor analyses of the WJ- R and KAIT in a non- white sample. Journal of School Psycholog y, 36, 151–182.
Flanagan, D. P., McGrew, K. S., & Ortiz, S. O. (2000). The Wechsler intelligence scales and
CHC theory: A contemporary approach to interpretation. Boston: Allyn & Bacon.
34 ESSENTIALS OF CROSS-BATTERY ASSESSMENT
Flanagan, D. P., & Ortiz, S. O. (2001). Essentials of cross- battery assessment. New York: Wiley.
Flanagan, D. P., Ortiz, S. O., Alfonso, V. C., & Mascolo, J. T. (2002). Achievement test
desk reference: Comprehensive assessment of learning disabilities. Boston: Allyn & Bacon.Flanagan, D. P., Ortiz, S. O., Alfonso, V. C., & Mascolo, J. T. (2006). Achievement test
desk reference: A guide to learning disability identifi cation (2nd ed.). New York: Wiley.Floyd, R. G., Bergeron, R., & Alfonso, V. C. (2006). Cattell- Horn- Carroll cognitive
ability profi les of poor comprehenders. Reading and Writing, 19(5), 427–456.Floyd, R. G., Keith, T. Z., Taub, G. E., & McGrew, K. S. (in press). Cattell- Horn-
Carroll cognitive abilities and their effects on reading decoding skills: g has indirect effects, more specifi c abilities have direct effects. School Psycholog y Quarterly.
Flynn, J. R. (1984). The mean IQ of Americans: Massive gains 1932 to 1978. Psychologi-
cal Bulletin, 95, 29–51.Glutting, J. J., Adams, W., & Sheslow, D. (2002). Wide Range Intelligence Test. Wilming-
ton, DE: Wide Range, Inc.Guilford, J. P. (1954). Psychometric methods (2nd ed.). New York: McGraw- Hill.Horn, J. L. (1991). Measurement of intellectual capabilities: A review of theory. In
K. S. McGrew, J. K. Werder, & R. W. Woodcock (Eds.), Woodcock- Johnson technical
manual (pp. 197–232). Chicago: Riverside Publishing.Kaufman, A. S. (2000). Foreword. In D. P. Flanagan, K. S. McGrew, & S. O. Ortiz
(Eds.), The Wechsler intelligence scales and Gf- Gc theory: A contemporary approach to interpre-
tation (pp. xiii–xv). Boston: Allyn & Bacon.Kaufman, A. S., & Kaufman, N. L. (1983). Kaufman Assessment Battery for Children.
Circle Pines, MN: American Guidance Service.Kaufman, A. S., & Kaufman, N. L. (1993). Kaufman Adolescent and Adult Intelligence Test.
Circle Pines, MN: American Guidance Service.Kaufman, A. S., & Kaufman, N. L. (2004). Kaufman Assessment Battery for Children–
Second Edition. Circle Pines, MN: AGS Publishing.Keith, T., Fine, J., Taub, G., Reynolds, M., and Kranzler, J. (2006). Higher order,
multiple- sample, confi rmatory factor analysis of the Wechsler Intelligence Scale for Children–Fourth Edition: What does it measure. School Psycholog y Quarterly, 35, 108–127.
Keith, T. Z., Kranzler, J., & Flanagan, D. P. (2001). What does the cognitive assess-ment system (CAS) measure? Conjoint confi rmatory factor analysis of the cognitive assessment system (CAS) and the Woodcock- Johnson tests (3rd ed.). School Psychol-
og y Review, 30(1), 89–119.Lezak, M. D. (1976). Neuropsychological assessment. New York: Oxford University Press.Lezak, M. D. (1995). Neuropsychological assessment (3rd ed.). New York: Oxford Univer-
sity Press.McGrew, K. S. (1994). Clinical interpretation of the Woodcock- Johnson Tests of Cognitive Abil-
ity–Revised. Boston: Allyn & Bacon.McGrew, K. S. (1997). Analysis of the major intelligence batteries according to a
proposed comprehensive Gf- Gc framework. In D. P. Flanagan, J. L. Genshaft, & P. L. Harrison (Eds.), Contemporary intellectual assessment: Theories, tests, and issues (pp. 151–180). New York: Guilford.
OVERVIEW 35
McGrew, K. S. (2005). The Cattell- Horn- Carroll theory of cognitive abilities: Past, present, and future. In D. P. Flanagan & P. L. Harrison (Eds.), Contemporary intel-
lectual assessment: Theories, tests, and issues (2nd ed., pp. 136–182). New York: Guil-ford.
McGrew, K. S., & Flanagan, D. P. (1998). The intelligence test desk reference (ITDR): Gf- Gc
cross- battery assessment. Boston: Allyn & Bacon.Messick, S. (1989). Validity. In R. Linn (Ed.), Educational Measurement (3rd ed., pp.
131–104). Washington, DC: American Council on Education.Messick, S. (1992). Multiple intelligences or multilevel intelligence? Selective emphasis
on distinctive properties of hierarchy: On Gardner’s Frames of Mind and Sternberg’s Beyond IQ in the context of theory and research on the structure of human abilities. Psychological Inquiry, 3(4), 365–384.
Messick, S. (1995). Validity of psychological assessment: Validation of inferences from persons’ responses and performances as scientifi c inquiry into score meaning. American Psychologist, 50, 741–749.
Ortiz, S. O., & Flanagan, D. P. (2002). Best practices in working with culturally and linguistically diverse children and families. In A. Thomas & J. Grimes (Eds.), Best
practices in school psycholog y IV (pp. 1351–1372). Washington, DC: National Associa-tion of School Psychologists.
Phelps, L., McGrew, K. S., Knopik, S. N., & Ford, L. (2005). The general (g), broad, and narrow CHC stratum characteristics of the WJ III and WISC- III tests: A con-fi rmatory cross- battery investigation. School Psycholog y Quarterly, 20(1), 51–65.
Reynolds, C. R., & Kamphaus, R. W. (2003). Reynolds Intellectual Assessment Scales. Lutz, FL: Psychological Assessment Resources.
Roid, G. H. (2003). Stanford- Binet Intelligence Scales–Fifth Edition. Itasca, IL: Riverside Publishing.
Sternberg, R. J., & Kaufman, J. C. (1998). Human abilities. Annual Review of Psycholog y,
49, 479–502.Thorndike, R. L., Hagen, E. P., & Sattler, J. M. (1986). Stanford- Binet Intelligence Scale–
Fourth Edition. Chicago: Riverside Publishing.Vanderwood, M. L., McGrew, K. S., Flanagan, D. P., & Keith, T. Z. (2002). The con-
tribution of general and specifi c cognitive abilities to reading achievement. Learning
and Individual Differences, 13, 159–188.Wechsler, D. (1981). Wechsler Adult Intelligence Scale–Revised. San Antonio, TX: The Psy-
chological Corporation.Wechsler, D. (1989). Wechsler Preschool and Primary Scale of Intelligence–Revised. San Anto-
nio, TX: The Psychological Corporation.Wechsler, D. (1991). Wechsler Intelligence Scale for Children–Revised. San Antonio, TX: The
Psychological Corporation.Wechsler, D. (1997). Wechsler Adult Intelligence Scale–Third Edition. San Antonio, TX: The
Psychological Corporation.Wechsler, D. (2002). Wechsler Preschool and Primary Scale of Intelligence–Third Edition. San
Antonio, TX: The Psychological Association.Wechsler, D. (2003). Wechsler Intelligence Scale for Children–Fourth Edition. San Antonio,
TX: The Psychological Association.
36 ESSENTIALS OF CROSS-BATTERY ASSESSMENT
Wilson, B. C. (1992). The neuropsychological assessment of the preschool child: A branching model. In I. Rapin & S. I. Segalowitz (Eds.), Handbook of neuropsycholog y:
Child neuropsycholog y (Vol. 6, pp. 377–394). Amsterdam: Elsevier.Woodcock, R. W. (1990). Theoretical foundations of the WJ- R measures of cognitive
ability. Journal of Psychoeducational Assessment, 8, 231–258.Woodcock, R. W., & Johnson, M. B. (1989). Woodcock- Johnson Psycho- Educational Bat-
tery–Revised. Chicago: Riverside Publishing.Woodcock, R. W., McGrew, K. S., & Mather, N. (2001). Woodcock- Johnson III Tests of
Cognitive Abilities. Itasca, IL: Riverside Publishing.Woodcock, R. W., McGrew, K. S., Mather, N., & Schrank, F. A. (2003). Diagnostic
supplement to the Woodcock- Johnson III Test of Cognitive Abilities. Itasca, IL: Riverside Publishing.
TEST YOURSELF
1. The XBA classifi cation system has had a positive impact on communi-cation among practitioners, has improved research on the relations be-tween cognitive and academic abilities, and has resulted in substantial improvements in the measurement of cognitive constructs, as seen in the design and structure of current intelligence batteries. True or False?
2. Fluid Intelligence (Gf ), Crystallized Intelligence (Gc), and Visual Pro-cessing (Gv) are examples of
(a) general (stratum III) ability.(b) broad (stratum II) abilities.(c) narrow (stratum I) abilities.(d) none of the above.
3. Two broad abilities not measured by many intelligence batteries pub-lished prior to 2000 that are now measured by the majority of intelli-gence batteries available today are
(a) Gc and Gv.(b) Gf and Ga.(c) Gf and Gsm.(d) Gsm and Gt.
4. The three pillars of the XBA approach are CHC theory, CHC broad (stratum II) classifi cations of cognitive and achievement tests, and
(a) CHC narrow (stratum I) classifi cations of cognitive and achievement tests.
(b) CHC general (stratum III) classifi cations of cognitive and achieve-ment tests.
(c) a and b.(d) neither a nor b.
S S
OVERVIEW 37
5. The second guiding principle of the XBA approach is to
(a) use as many intelligence batteries as necessary to answer the refer-ral concerns.
(b) use subtests and clusters from a single battery whenever possible to represent broad CHC abilities / processes.
(c) select tests that have been classifi ed through an acceptable method, such as through CHC theory- driven factor analyses or expert con-sensus content- validity studies.
(d) create broad CHC clusters instead of narrow CHC clusters when possible.
6. An example of a cluster that contains construct- irrelevant variance is the
(a) WISC- IV VCI.(b) WJ III Comprehension- Knowledge Factor.(c) WAIS- III VIQ.(d) KABC- II Simultaneous / Gv Scale.
7. Most clusters that are found in today’s comprehensive intelligence bat-teries are both relatively pure (i.e., containing only construct- relevant tests) and well represented (i.e., containing qualitatively different mea-sures of the broad ability / process represented by the cluster). True or False?
8. Which of the following is not a good descriptor of the XBA approach?
(a) Time- effi cient(b) Theory- focused(c) Test kit–focused(d) Empirically supported
9. All of the following narrow abilities / processes fall under Gc except
(a) Listening Ability (LS).(b) Language Development (LD).(c) Lexical Knowledge (VL).(d) English Usage Knowledge (EU).
10. When conducting XBA, it is important to select tests from a limited number of batteries. True or False?