|
|
|
|
|
HOME | BIBLIOGRAPHY
| TRAININGS | DOWNLOADS | BIOGRAPHIES
| LINKS | FAQS FAQ Click
on one of the four topics below that most closely matches your question
and you'll be taken there automatically where you can scroll down
the section to find the information which may address your
questions or problems:
General Questions Test Interpretation Graphs Culture-Language Interpretive Matrix Errata
The previous
cross-battery worksheets supplied in The
Intelligence Test Desk Reference had "ratings" of how good a measure
the different subtests were of each narrow ability area. I notice in
the
current cross battery assessment CD, the AREAS are rated as to how well
they
relate to different areas of achievement, but I don't see worksheets
that show
which subtests from various instruments are the best measures of the
relevant
narrow ability. Is there an updated worksheet available so that
the
"best" subtests to measure each narrow ability can be seen and
chosen? I am unaware of worksheets that
indicate the "best" subtests that measure each narrow ability since
most factor analyses have not been conducted at the narrow ability
level.
However, expert consensus studies have been conducted at the broad and
narrow
ability levels. Appendix B that begins on page
298 in the Essentials of Cross-Battery Assessment, Second Edition
indicates
what subtests are strong, moderate, or mixed measures of CHC abilities.
In
addition, Rapid References 2.2 - 2.4 on pages 55-61 may be helpful. Will the program
automatically compute an average for
the broad G’s if the two subtests are measuring the same thing
(are not
qualitatively different)?
If a cluster is interpretable, then
it will be calculated by the program. You need to determine if the
cluster is
broad or narrow based on the classifications of the tests that comprise
it.
That is, if the subtests comprising a cluster measure the same narrow
ability,
then the cluster generated by the program represents a narrow ability. When using the WJ
III COG in which you get a score for all 7 Gs, do you use those scores
or do
you use the scores from the XBA chart which you get from the disc in
your book? Many times the two
scores are very similar. Other times the scores are different. Example:
Gc
score on COG computer print out is 84 (below average) Gc score from XBA
disc is
88 (average) (Memory for words is 86 & Numbers Reversed is
89). So, do you only
use the XBA chart scores; can you pick & choose, or only use
computer
printout scores?
When you give the WJ III COG, then you are suppose
to
enter the scores in the WJ III tab. Only when the WJ III CHC Clusters
are not
interpretable do you consider using the CHC tab for WJ III tests. So, for example, I administer WJ III AS and
CF. AS = 80 and CF = 105.
In this instance, WJ III Compu Score will
give you a Gf cluster, but the WJ III
tab of the XBA DMIA will say that Gf
is not interpretable. The examiner would then want to follow up with
another
measure of general sequential reasoning b/c AS was in the deficient
range.
After following up, the examiner will now have 3 scores for Gf
(2 that measured General Sequential
Reasoning and one that measured Induction). Because there are no actual
norms
for this combination of three subtests, they should be entered into the
CHC tab
in the Gf section. Depending on where
the 3 subtest score falls, the program will either indicate that no
cluster can
be calculated or will calculate a cluster based on two subtests and
report the
3 subtest as an outlier. So, if the third subtest was a standard score
of 82,
for example, then the CHC tab would show a cluster of 81 and an outlier
of 105,
meaning that in the area of Gf, the
student demonstrates average ability to reason inductively but has a
deficit in
his or her ability to reason deductively. So, as a general rule, you
always use
actual norms when they are available and when the cluster (scale,
index,
composite) is interpretable. If actual norms are not available, then
the CHC
tab is used to determine whether the scores can be aggregated. Note that there are times when a cluster may
be reported as not interpretable on the test tab and interpretable on
the CHC
tab. For example, if AS was a 65 and CF was an 82, then the WJ III tab
would
state that Gf is not interpretable.
In this case, since both scores are deficient, we wouldn't necessarily
need to
follow up. As such, if these scores were entered into the CHC tab, a
cluster
would be reported. A Gf cluster of 74
would be reported b/c both scores (65 and 82) are in the same normative
range
(i.e., the deficient range). On the CHC tab, is
there a way to type in CTOPP rapid
naming tests under the Glr cluster...so that the names of these tests
appear on
the CHC graph? No. You can enter the score in the Glr
section without selecting a subtest name and then print the
graph and write the subtest name(s) in by hand. Alternatively, if you
used the
CTOPP to supplement the WISC-IV, for example, you can type in CTOPP
information
at the end of the WISC-IV tab under section “other clusters to be
plotted.” I
cannot clear data by clicking on the button, nor does the program
indicate a
Yes or No for the Is the Composite Interpretable question. What can I
do? Be sure to enable macros when
you start the program. I'm trying to learn how to use the DMIA for
cross battery
assessment, and I haven't been able to save data. I followed the
directions
included in the "notes" and clicked on "Save As" and
renamed the file. I get a pop-up saying it is a read-only file. Since
cross
battery assessment involves a lot of testing, I don't get it all done
in one
testing session. How can I put in data as I get it instead of having to
wait until I
am finished testing to use the DMIA? How do I save data before moving
to the
next case? The reason for the read-only
pop-up is because the default location for saving the newly named file
is the
CD itself. The location of the new file obviously can not be on the CD
because
it is protected. Therefore, the location where the file is to be saved
needs to
be changed to either another location on the hard drive (e.g., My
Documents) or
a removable drive (e,g., flash drive).
After the new file is saved for the first
time in a new location, then data can be entered as it is available and
there
is no need to do a "Save As"; "Save" is sufficient. To
re-open the file, there is no need to run the CD, the person can run
the file
by simply double-clicking that renamed
file. To open a clean empty file for another
case, the CD should be run and the "Save as" procedure should be
followed again. When I enter the standard scores, it gives me
the
percentiles. However, it doesn't tell me the cluster scores, the
descriptive
categories or whether the cluster is interpretable. Here's what
I've already done. I've set my security setting
to "low" to use it. Also, I've saved it under a different name. I
don't know whether this has anything to do with it, but when I
initially go into
the workbook, I get a message that says, "Workbook is protected and
cannot
be changed." I've tried to unprotect it, but it tells me I need a
password
to do that. First, the protected workbook
message shows up any time there is an attempt to click on a cell that
is
automated and not available for user input. Users may only enter data
in
"unprotected" cells. There is no need to alter protected cells as
they are there for information or for automation and will display
information
as becomes necessary depending on the user data that are entered. The
protection is necessary to protect the program source code from being
altered. Second, cluster and composite
scores are only calculated automatically if scores are entered on the
CHC tab.
They are not automatically calculated by the program for specific
cognitive
batteries, that is, for data entered on battery-specific tabs (e.g.,
WISC-IV).
We do this for several reasons. One the program is not intended to be a
scoring
program for other tests. It is designed more to assist in the use of
data
gathered from tests in a cross-battery and theory-based measurement and
interpretation manner. In this way, it does not prevent users from
using the
data they have already generated from scoring a particular test and it
has the
advantage of allowing the user to use the actual cluster or composite
scores
that are derived from the norms of the test given. When a specific
battery is
used, it may be helpful to maintain the composites or clusters as they
were
derived from the actual norms of the test. So for any tab other than
the CHC
tab, it is necessary for the user to fully score the battery given
before using
the XBA-DMIA and then enter both the subtest and
cluster/composite/IQ/Index
scores into the appropriate tab in the program. If, on the other hand
you are
doing cross-battery assessment, you do have the option of entering the
scores
on the CHC tab. Here you would only need to enter the subtest scores
and the
program will automatically convert them to a common metric and will
automatically
calculate a cluster/composite score. It will also automatically report
the
descriptive category, and will automatically report whether the
composite is
interpretable or not. On the test-specific tabs, the
program operates differently because it needs to compare the
cluster/composite
score you enter with the subtest scores that comprise it before it can
make a
determination regarding interpretability. So, if you enter the
composite score
first, you will get the descriptive category, but it will not indicate
whether
it is interpretable. If you then enter the subtest scores that comprise
that
cluster/composite (at least two but sometimes three subtests are needed
such as
VCI on WISC-IV), it will then indicate whether the cluster is or is not
interpretable.
The same will happen with the FSIQ and GAI and other such scores. Since
they
are made up of a combination of cluster/composite scores, the
interpretability
will not be reported until the appropriate number of cluster/composite
scores
are entered. This works in reverse too, that is, if you enter the
subtest
scores that comprise a particular cluster/composite, the descriptive
category
for it and interpretability will not show up until the actual score for
the
cluster/composite is entered. Again, the program must wait for the all
the
information to be entered before it can determine interpretability. WINDOWS
There are some
Windows Vista issues that may affect the use
of the automated programs on the CD Rom. First,
the program opens with the window
absolutely minimized on my screen, maybe on yours too. I have to hit
the
maximize button to get it to open up, and then it's full screen instead
of a
window like normal. The publisher's programmers will need to look at
this issue
because it is annoying and a real pain when you first start the
program.
Second, Why can’t I enter data for the KABC-II
Story Completion and
Pattern Reasoning subtests for the Planning/Gf scale? The little red triangle in the
upper right corner is meant to instruct the user to enter the data for
Story
Completion and Pattern Reasoning under the Simultaneous/Gv
section and they will be automatically carried over to the
Planning/Gf section. Thus, you cannot
enter scores in Planning/Gf directly,
you must enter them in Simultaneous/Gv
and allow the program to carry the scores over automatically. If you administered all of the DAS-II you would
want to put
in four subtest scores in the Gc section of the CHC tab, but there is
only room
for three. Is there a way to add another cell under Gc or any other
area? The basic answer, unfortunately,
is that there is no way to include four scores for any broad ability on
the CHC
tab/graph. This is perhaps something we can do for an update, allow
maybe four
or more scores to be entered on that tab. However, if as the person
indicates,
the entire DAS-II was administered, then there's really no reason to
use the
CHC tab in the first place. The scores should rightly be entered in the
DAS-II
tab (the last tab on the right side) where all four scores for Gc can
be
entered, in addition to all other scores that may have been obtained.
The DAS-II
tab will then graph all the scores and will even graph six additional
scores
that can be entered at the very bottom in the blank spaces, should
there be any
supplemental testing or scores from other batteries. The CHC broad
abilities
are noted on the DAS-II tab (e.g., Gc, Gf, Gv, Gs, Gsm).
In your Rapid Reference Tables 2.2 to 2.4, you
list
subtests that measure CHC narrow abilities/processes that relate
significantly
to achievement in several academic areas. How have you determined that
relationship? The subtests listed in Table 2.2 to 2.4 are measures of the CHC abilities/processes that in turn predict the respective areas of achievement. There are both content and construct validity evidence for the subtests listed in these tables as measures of the different CHC abilities/processes. If we know that there is a strong relationship between phonological processing and reading, and if we want to measure phonological processing in a child who is having reading difficulties, then we would want to select tests that are valid measures of phonological processing. These tests are printed in Tables 2.2 to 2.4.
Question: Should
we be as concerned about interpreting achievement subtest scores within
a
cluster with statistically significant differences as we are about
interpreting
cognitive ability subtest scores within a cluster with statistically
significant differences? Question: I gave
the Glr cluster for the WJ III. When I entered the scores in under the
WJ tab
they came back as non interpretable Visual-Auditory Learning was 84 and
Retrieval Fluency was 63. When I entered these scores under the CHC tab
it gave
me an average of 74. The WJ scorer gave me a Glr cluster of 71, not 74.
Why was
it considered interpretable under the CHC tab, why were the cluster
scores off
by three points? Question: Right
now we have 3 qualitatively different narrow abilities entered under
Glr. In the software it
looks as follows: LONG TERM
RETRIEVAL WJ III VAL = 99 WJ III Retrieval
Fluency = 111 WJ III VAL-Del =
82 (outlier) Composite according
to the software = 105 Other Glr tests
administered: WJ III Memory for Names = 104 WJ III Memory for
Names Delayed = 113 How is the
composite getting 105? How is this
interpretable given the difference? We only recommend using the
CHC tab, when actual norm-based clusters are not interpretable. If you use the WJ III tab, you will see that
the Glr cluster based on VAL and RF
is interpretable. Therefore, the WJ III Glr cluster should be reported and
interpreted. The 105 reported on the CHC tab is an arithmetic average
of VAL
and RF only and is within the SEM of the WJ III Glr
cluster based on actual norms. Note that the program will
calculate an arithmetic average for scores that fall within the same
normative
range, as was the case with your VAL and RF scores.
All the rules for cluster calculation and
interpretation are clearly spelled out in the book (including why the
WJ III
VAL-Del was reported as an outlier; see Chapter 3).
Question: What
happens to an outlier when the program is computing a cluster? For
example, SHORT-TERM MEMORY WJ III Memory for
Words = 79
(outlier) WJ III Numbers
Rev = 98 WJ III Auditory
Working Memory = 105 Composite for on
CB2 software = 102 So it seems like
the program is throwing out the 79 to get the composite. Is this true?
Also
after working across the XBA2 book and ATDR2, it seems there are some
slight
variations on how to handle things. Why is this? Note that the differences
between the ATDR-2 worksheets and the XBA DMIA (program on the CD-ROM)
are
substantial. The main difference is that the XBA DMIA tells you what
clusters
are and are not interpretable and bases cluster interpretation (2 or 3
subtest
combinations) on a detailed, psychometrically and theoretically
defensible set
of rules. When using the ATDR-2 worksheets, you need to know more about
how to
interpret the results.... the XBA DMIA does more for you...it
simplifies the
process.
Question: We are interested in
seeing from an interpretation viewpoint how you would deal with student
profiles where the performance is expected for the child's diverse
background
and where it is not expected. Validity statement where declining
pattern is evident: Because the student is not a native
English speaker, it is necessary to establish the validity of the
results
obtained from testing to ensure that they are accurate estimates of
ability or
knowledge and not the manifestation of cultural or linguistic
differences. To
this end, a systematic evaluation of the possible effects of lack of
acculturation and limited English proficiency was carried out via use
of the
Culture-Language Interpretive Matrix (C-LIM). A careful review of the student’s
test data as entered into the C-LIM reveals a pattern of decline that
is
typical of and within the range that would be expected of other
individuals
with similar cultural and linguistic backgrounds. This declining
pattern of
test performance suggests that test performance was due primarily to
the
influence of cultural and linguistic factors rather than actual
ability.
Accordingly, the test results evaluated here cannot be considered valid
and are
not interpretable from a strict psychometric standpoint. However, given
that
the observed pattern is consistent with performance that is typical of
non-disabled, culturally and linguistically diverse individuals with
average
ability, it can be reasonably concluded that this student’s
abilities are also
within the average range of performance and does not suggest or support
the
presence of any type of disability. Validity statement where declining
pattern is NOT evident: Because the student is not a native
English speaker, it is necessary to establish the validity of the
results
obtained from testing to ensure that they are accurate estimates of
ability or
knowledge and not the manifestation of cultural or linguistic
differences. To
this end, a systematic evaluation of the possible effects of lack of
acculturation and limited English proficiency was carried out via use
of the
Culture-Language Interpretive Matrix (C-LIM). A careful review of the student’s test
data as entered
into the C-LIM does not appear to reveal a pattern of decline that is
typical
of or within the range that would be expected of other individuals with
similar
cultural and linguistic backgrounds. The pattern of test performance
does not
decline systematically and suggests that test performance was not due
primarily
to the influence of cultural and linguistic factors. Although such
influences
may be contributory factors, they are not believed to be the primary
reasons
for the observed pattern. In addition, other extraneous factors that
might
account for the observed pattern (for example, lack of motivation,
fatigue, incorrect
administration/scoring, emotional/behavioral problems) have been
excluded. This
indicates that the test results can be considered valid, interpretable,
and are
likely to be good estimates of the student’s actual ability or
knowledge.
Because the observed pattern is not consistent with performance that is
typical
of non-disabled, culturally and linguistically diverse individuals with
average
ability, it can be reasonably concluded, if supported by additional
data, that
the student’s test performance may be attributed primarily to the
presence of a
learning disability. Note that you may substitute the
name of the student who is being evaluated in place of where I have
written
"the student's" to make the report more personalized. Also, if you
use the latter statement, that the results are valid, you would then
proceed to
using the XBA DMIA and provide an analysis and interpretation of those
results
as they might pertain to the presence of a learning disability just as
you
would for any other student. These and additional interpretive
statements for
use with ELLs can be found on the Downloads tab on this website. Question: I administered the WISC-IV and
administered additional subtests from the WJ III to get the Gv and Glr
cluster.
The Gv cluster was based on the aggregate performance on block design
and
picture recognition. Can I report the average standard score from the
cross
battery as the Gv cluster since the XBA score is only the average of
the two
subtests? Or should I give the spatial
relation from the WJ III to get the cluster score for Gv processing? I guess what I am asking is if we can get a
cluster score when we use subtests from two different batteries using
the XBA. Question: For my
bilingual kids, I generally administer the WJ III and get all 7
cognitive
clusters. When the broad cluster is nonunitary, I administer a 3rd
measure of
the weaker narrow ability. Let's say these two measures of the weaker
ability
do not converge and the higher score is from the 2nd assessment battery
(i.e.,
WISC-IV). Then the student's cognitive ability (i.e., Gf) is an
aggregate of
the one subtest of the WJ III and one subtest of the WISC-IV. Given
this, can I
report the average score based on XBA as the “broad
cluster score” for the cognitive ability (i.e., Gf)?
It looks like the graph is about 5 points off, is that correct? Because of the “3-D” nature of the graphs, it appears that the shading itself is 5 points off AND the normative range labels at the top are also exactly 5 points off which makes it seem the conversions are off, but in fact they are not. For example, if the cluster score for the Sequential/Gsm is 85, on the graphs, the bar appears to be graphed starting at 80 rather than 85. Similarly, the “normative range” label and indicator line also appears to be shifted by 5 points. In addition, the range of the bar that is calculated is also correct (i.e., SEM of +/- 5 points), but because of the 3-D effect of the bars, which makes them "float above the lines," it is easy to visually misinterpret which scores are represented.
The average range on the CHC Broad Ability Data
chart, when
printed, is off and shown as 87 to 117 instead of 85 to 115. On the screen, the shading is
fine as are the descriptive range brackets. When you print the page,
however,
Excel shifts the brackets and shaded range over and down slightly,
altering the
actual average range. This is an artifact of the source code in Excel
and a
limitation of the program that we were unable to overcome. The
problem is that the gray shading and the descriptive range labels and
even the
brackets that delineate the areas are NOT an actual part of the graph.
They are
separate graphical "objects" that we add on top of the graph that is
generated automatically. what happens is that when the program goes to
prepare
the document for printing, it imposes the constraints of the printer
that will
be used. For example, most inkjet printers can only print to about 1/2"
from the bottom and top of the page whereas most laser printers can go
to
1/4" top and bottom. So the program rescales the document to fit the
printer's limitations. Generally, everything would be adjusted
accordingly. But
in our case, because the graph is generated as a whole, and is not
connected to
the descriptive range labels, shading, and such, it may, at times and
with
certain printers, stretch or shrink the graph one way or another to fit
it on
the paper as it feels is best. We tried to limit this
shrinking/stretching as
much as possible by having the graph's vertical range remain constant.
But the
program nevertheless "adjusts" the graph horizontally any time it
must adjust it vertically. Remember, the graphs are printed in
landscape, so on
some printers, a half inch from the top and bottom, or even a half inch
on the
bottom and a quarter inch on top, means that 3/4" to 1" of the paper
cannot be used. That limits the graph to 7 1/2" to 7 3/4" in vertical
height in landscape mode. Horizontally, it still leaves as much as 10"
to
10 1/4" which is more than enough to fit the proper aspect ratio of the
graph if it has to be shrunk vertically. Thus, on the page, the
"objects" we impose upon the graph do not not need to move at all and
remain where they are. However, if any single element of the graph
needs to be
readjusted (that is, scaled down vertically), then all elements created
along
with the graph, being a part of the graph, are likewise moved resulting
in a
corresponding horizontal adjustment of the graph. So the graph changes
slightly
but the labels do not and the printing looks off. The gray shading
looks ok on
screen because there is no margin constraint in the way the graph or
objects
may be displayed on a page. It becomes a problem when the graph is to
be
printed on a printer that cannot match the exact layout of the graph.
This is a
problem, as I mentioned in my prior response, with the way that Excel
handles
documents for printing. Word does the same thing in many cases. It also
has to
do with the way in which graphs are created and what you can do with
them. MS
never really thought anyone might superimpose graphics on top of the
graph and
thought that all the features available in the graph's creation were
enough so
that it didn't worry about having to adjust the graph for printing and
the
problems it might create, thinking there would not be anything else to
worry
about, or at least nothing that required the precision we do. Bottom
line is
that there are two possible solutions I can think of. One is that we
try, and
this may be very difficult, to make the graphs smaller. Yes, smaller.
They will
be less readable on a regular computer screen and maybe even when
printed out,
but if we can make the graphs fit on the page with about a 1/2" margin
all
the way around, then we will be within the parameters of 99% of all
current
printers and Excel would not then need to adjust the graph and the
shading and
labels wouldn't move. We could tell people to use a more modern
printer, as
some new printers have edge-to-edge capability, but people may not love
that
option because it means more money (and only about 10% of all
printers--mostly
photo printers--have edge-to-edge compatibility), or they may not even
know how
to configure their printers for margin-less printing since that option
usually
assumes photo printing which is not what is being done here. The second
hope we
have is that the newest version of Excel (2007) might have more
features
available so that we can apply the "objects" as a part of the graph,
rather than external to it or superimposed on it. If we can do this,
the graph
can adjust as necessary and Excel will adjust our labels right along
with it. There does not seem
to be a feature on the CD program which
automatically combines all the entered information, for example, WISC +
CHC
scores - and displays it all on one table/graph. Is that correct? Note that below all entered
scores on each test tab, there is a place for entering data from other
batteries and tests. You may enter up to 6 clusters. For example, if
you
administered the WISC-IV and entered all WISC-IV data on the WISC-IV
tab, but
also gave the WJ III Ga and Glr clusters to supplement the WISC-IV,
then the Ga
and Glr clusters can be entered into the WISC-IV tab and the WISC-IV
graph
would display these clusters along with the WISC-IV data.
I like to add some achievement clusters here
as well so that I can see the pattern of strengths and weaknesses
across both
academic and cognitive areas on the corresponding graph.
Is it possible to cut
and paste a student's completed
DAS-II graph from the excel software into a MS Word report? Yes, you can cut and paste the graph to a report
in
Word. To do so, click in the blank space just above the title of the
graph.
You'll notice that small black squares will appear in the corners and
along the
midpoint of the graph. This indicates that the entire graph has been
selected.
If you see only a portion of the graph selected, click in the space
outside the
graph and the squares will disappear and you can try selecting again.
Once you
have the graph selected, simply click on the "copy" icon at the top
of the menu bar (between the cut and paste icons), or else click on the
Edit
command, and then select the "copy" option. Note, by the way, that
you cannot use the "right mouse click, then copy" option here because
it is not allowed in this case when using Excel. Once the graph has
been
copied, simply open up your word document and select the paste icon
from the
menu bar or again choose the Edit command and then the paste option.
Alternatively, you can now use the shortcut where you simply right
click
anywhere with the mouse then choose the paste option which will be in
the pop
up menu. The entire graph will then be pasted. This works best if your
Word
document is in landscape mode, but it does work fine in portrait mode.
Note
also that the grey shaded area may shift a little when you paste the
document.
This is not anything we can control at the moment given the limitations
in both
Word and Excel. If you are familiar with editing objects in Word, you
can
select the shaded area and move it to more closely align with the
values. But
if you do not know how to do this, it's best just to leave it alone. Can the DAS-II
t-scores from the DAS-II page be transferred
directly into the CHC table? No. The DAS-II scores on the DAS-II tab are automatically converted to the deviation IQ metric (X = 100, SD = 15). These converted scores are the ones that should be transferred to the DAS-II tab.
CULTURE-LANGUAGE
INTERPRETIVE MATRIX Can
input scores from the Bateria III COG (instead of the WJ III COG) into
the
Cross-Battery software. Our District serves a large bilingual
population and we
often use the Bateria III COG. We are frequently faced with scenarios
that
would benefit from the cross-battery approach. However the book just
refers to
the WJ III and we fear it would not be valid to use the software with
the
Bateria III COG. The short answer is, yes, you can use the
standard
scores you derive for each subtest on the Bateria-III in the XBA DMIA
in the
corresponding data entry cells on the WJ III tab. The tests are
structured in
parallel fashion, so there's no problem in knowing which test is which
or where
it goes in the program. You may also enter the cluster scores from the
Bateria-III into the program as well and it will inform you if the
cluster is
or is not interpretable. You can also enter any cluster or subtests in
the
blank spaces provided at the bottom of the WJ III sheet to include
additional
subtests from other batteries or the WJ III Diagnostic Supplement. The
main
reason you can enter the scores for the Bateria-III, apart from the
parallel
structure of the tests, is that you are entering standard scores in the
program, not raw scores. Our program is not a scoring program, but only
a data
management and interpretive assistant. Thus, it requires the use of
standard
scores only. This has several advantages. First, the standard scores,
whether
from the WJ III or the Bateria-III, are calculated by the computer
scoring
program, not by us, and thus, each standard score is derived from the
appropriate norming sample for each test. So each score is derived
correctly
and in accordance with the procedures of each test and there is no
alteration of
the score whatsoever. Second, standard scores are readily
interchangeable.
Whether you enter scaled scores (such as with the Wechslers and
KABC-II),
T-scores (such as with the DAS), or deviation IQ scores (such as with
the WJIII
and Bateria-III), they all share the same properties except for the
scaling
which is arbitrary. For example, a scaled score of 7 on the WISC is the
same as
a T-score of 40 on the DAS-II which is the same as an 85 on the
WJIII/Bateria-III. Third, although they are different metrics, the
percentile
rank is the same for all of them (16th PR). It does not change. And
fourth,
because the percentile ranks are equivalent, the scores are all
interpreted by
our program consistently (in this case, as being at the very bottom end
of the
average range). So there is no significant issue in using standard
scores from
the Bateria-III in the program. In the next version, we'll likely
include a
separate tab for the Bateria-III where people can feel comfortable
putting the
scores in a place for which they are specifically designed. I
am assuming that the C-LIM is used with English language learners that
are
dominant in English because the test tabs are tests with English
instructions
mostly. I have been getting many referrals for testing on children who
are
still Spanish dominant in which case the cognitive instruments that I
use may
include subtests from the WISC-4 Spanish and the Bateria III Pruebas de
habilidades cognitivas (sometimes I use subtests from the UNIT of the
K-ABC
especially if I have non-unitary factors and I try to limit my testing
to 2
instruments) My questions are: 1.
Can the C-LIM possibly be used with the
above-mentioned cognitive instruments that are administered in Spanish? Yes, the C-LIM may be used with scores obtained
from the
Bateria-III and WISC-IV Spanish 2.
If yes, I am assuming that I would plug the scores
in to the general tab because I used cross-battery tests, correct? You may plug the scores into either the
test-specific
tabs or the general tab, whichever best suits the evaluation you are
doing. 3.
Would I plug in the scores for the Bateria
III #6 Pareo Visual to the WJ III #6 Visual matching tab? Yes, you can put the score for the subtest from
the
Bateria-III/WISC-IV Spanish that corresponds directly to the same test
on the
WJ III/WISC-IV Spanish You also mention the concept of dominance in
your
question and this has nothing to do, really, with whether the C-LIM is
or is
not applicable. Language dominance only tells you which language is
better developed
or sometimes preferred, but it does not tell you how proficient the
individual
is in either language. You could be dominant in one language but still
have
very low proficiency in it. Conversely you could have a very high level
of
proficiency in your dominant language and just a high level in the
other, which
would be great. But in the schools we run across kids who are often low
in
both, so dominance is not really a useful concept and it will not
affect the
type of pattern that should result in the C-LIM for the reasons which
I'll
explain below. And last, you should not be embarrassed at all
by your
questions. They are excellent questions and I got many of the same ones
in my
last presentation. In truth, what you are asking is really something
that comes
from an advanced understanding and use of the C-LIM so you should feel
proud
that you are actually way ahead of the curve, not behind it! As I said,
the
short answer to your first question is, yes, you can use the standard
scores
you derive for each subtests on the Bateria-III/WISC-IV Spanish in the
XBA DMIA
in the corresponding data entry cells on the WJ III/WISC-IV tabs. I
don't
recall exactly if the WISC-IV Spanish is or isn't parallel to the
English
version, but I believe it is, and I do know that the WJIII and
Bateria-III
tests are in fact structured in parallel fashion, so there should be no
problem
in knowing which test is which or where it goes in the program. Thus,
you may
also enter the cluster scores from the Bateria-III/WISC-IV into the
program as
well and it will inform you if the cluster is or is not interpretable.
You can
also enter any cluster or subtests in the blank spaces provided at the
bottom
of the tabs to include additional subtests from other batteries (UNIT,
KABC-II,
WJ III Diagnostic Supplement). The main reason you can enter the scores for
these
Spanish tests, apart from the parallel structure, is that you are
entering
standard scores in the program, not raw scores. Our program is not a
scoring
program, but only a data management and interpretive assistant. Thus,
it
requires the use of standard scores only. This has several advantages.
First,
the standard scores, whether from the WJ III or the Bateria-III, are
calculated
by the computer scoring program, not by us, and thus, each standard
score is
derived from the appropriate norming sample for each test. So each
score is
derived correctly and in accordance with the procedures of each test
and there
is no alteration of the score whatsoever. Second, standard scores are
readily
interchangeable. Whether you enter scaled scores (such as with the
Wechslers
and KABC-II), T-scores (such as with the DAS), or deviation IQ scores
(such as
with the WJ III and Bateria-III), they all share the same properties
except for
the scaling which is arbitrary. For example, a scaled score of 7 on the
WISC/WISC Spanish is the same as a T-score of 40 on the DAS-II which is
the
same as an 85 on the WJIII/Bateria-III. Third, although they are
different
metrics, the percentile rank is the same for all of them (16th PR). It
does not
change. And fourth, because the percentile ranks are equivalent, the
scores are
all interpreted by our program consistently (in this case, as being at
the very
bottom end of the average range). So there is no significant issue in
using
standard scores from the Bateria-III in the program. In the next
version, we'll
likely include a separate tab for the Bateria-III where people can feel
comfortable putting the scores in a place for which they are
specifically
designed. You can also use the scores from the
Bateria-III/WISC-IV
Spanish, actually, the clusters for the seven CHC abilities, in the SLD
Assistant which is used to evaluate the criterion for "an otherwise
normal
ability profile." If you are using the WISC-IV Spanish, however, you
will
need to do supplemental testing to ensure that all seven CHC domains
are
evaluated. In that program, you merely enter "yes" or "no"
for each of the abilities with respect to whether the cluster is intact
(within
the average range) or not. If valid clusters are reported by the XBA
DMIA and
if they are also classified as average, then you can answer "yes" and
if they are below average you can answer "no." Ok, all this aside, let me add some caveats
before you
do anything. If you are using the Bateria-III/WISC-IV Spanish, it means
you are
working with an individual who is not a native English speaker. Thus,
BEFORE
you would even consider evaluating you must first use the C-LIM to
determine
whether your results are valid or not. In general, the C-LIM was
designed to
evaluate the performance of bilinguals on tests administered to them in
English. This is because there is a very long history and body of
research
regarding how bilinguals perform on tests given to them in English. In
contrast
there is virtually no research on tests given to bilinguals and what
their
performance might be when the test is given in the native language, not
in
English. However, because the Bateria-III/WISC-IV Spanish, as with all
current
tests designed for speakers of languages other than English, are
actually
normed and were adapted primarily on monolingual speakers of Spanish,
living in
monolingual Spanish speaking countries, being raised by monolingual
Spanish
speaking parents, and being educated monolingually in Spanish, their
use with
bilinguals living in the U.S. is probably as problematic as it is with
any
other test normed on monolingual English speakers. Although both the
Bateria-III and WISC-IV Spanish claim to include "bilinguals" they
are not represented in the norms in any systematic way. That is,
language
proficiency in both languages and their respective levels of
acculturation are
not variables that are stratified or sampled randomly from the
population. The
norm samples remain primarily a monolingual group, whether English or
Spanish
speaking. Thus, it is likely, and this is only speculation at this
time, but I
do believe there is evidence to support it, that bilinguals will
exhibit the
same type of declining pattern of performance on the Bateria-III and
WISC-IV as
they do on the WJ III and WISC-IV or other such cognitive tests given
in
English. Therefore, you should not take your scores directly from the
Bateria-III or WISC-IV Spanish and place them in the XBA DMIA until you
have
first determined their validity with the C-LIM, just as you would with
any test
you may have given in English. We actually have some research going on
at the
moment that will examine the performance of bilinguals on the
Bateria-III, so
we may later be able to provide more guidance on the nature of the
declining
pattern. For now, I'd look at the low culture/low language tests and
see if
they are highest, the hi culture/hi language tests to see if they are
the
lowest, and the moderate culture/moderate language tests to make sure
they are
between the two. I wouldn't worry too much about the pattern of the
tests that
are off the diagonal, so to speak, as there may be changes in those
classifications in the future. In any event, the C-LIM will allow you
to
evaluate whether there is or is not a declining pattern of performance
and this
determination is crucial to whether you can have confidence that your
scores
are valid or not. In short, if the pattern declines systematically,
then the
results are not valid and you should not enter them in the XBA DMIA and
you
should report that they are invalid and thus uninterpretable, however,
that
would also mean that there is nothing wrong with the individual. If you
wish,
you are welcome to the following interpretive statements to help
explain your
findings and support your work. Validity statement where declining pattern is
evident: Because the student is not a native English
speaker, it
is necessary to establish the validity of the results obtained from
testing to
ensure that they are accurate estimates of ability or knowledge and not
the
manifestation of cultural or linguistic differences. To this end, a
systematic
evaluation of the possible effects of lack of acculturation and limited
English
proficiency was carried out via use of the Culture-Language
Interpretive Matrix
(C-LIM). A careful review of the student’s test data as entered
into the C-LIM
reveals a pattern of decline that is typical of and within the range
that would
be expected of other individuals with similar cultural and linguistic
backgrounds. This declining pattern of test performance suggests that
test
performance was due primarily to the influence of cultural and
linguistic
factors rather than actual ability. Accordingly, the test results
evaluated
here cannot be considered valid and are not interpretable from a strict
psychometric standpoint. However, given that the observed pattern is
consistent
with performance that is typical of non-disabled, culturally and
linguistically
diverse individuals with average ability, it can be reasonably
concluded that
this student’s abilities are also within the average range of
performance and
does not suggest or support the presence of any type of disability. Validity statement where declining pattern is
NOT
evident: Because the student is not a native English
speaker, it
is necessary to establish the validity of the results obtained from
testing to
ensure that they are accurate estimates of ability or knowledge and not
the
manifestation of cultural or linguistic differences. To this end, a
systematic
evaluation of the possible effects of lack of acculturation and limited
English
proficiency was carried out via use of the Culture-Language
Interpretive Matrix
(C-LIM). A careful review of the student’s test
data as entered
into the C-LIM does not appear to reveal a pattern of decline that is
typical
of or within the range that would be expected of other individuals with
similar
cultural and linguistic backgrounds. The pattern of test performance
does not
decline systematically and suggests that test performance was not due
primarily
to the influence of cultural and linguistic factors. Although such
influences
may be contributory factors, they are not believed to be the primary
reasons
for the observed pattern. In addition, other extraneous factors that
might
account for the observed pattern (for example, lack of motivation,
fatigue,
incorrect administration/scoring, emotional/behavioral problems) have
been
excluded. This indicates that the test results can be considered valid,
interpretable, and are likely to be good estimates of the
student’s actual
ability or knowledge. Because the observed pattern is not consistent
with
performance that is typical of non-disabled, culturally and
linguistically
diverse individuals with average ability, it can be reasonably
concluded, if
supported by additional data, that the student’s test performance
may be
attributed primarily to the presence of a learning disability. If you use the latter statement, that the results are valid, you would then proceed to using the XBA DMIA and provide an analysis of those results as they might pertain to the presence of a learning disability.
"...the initial step in evaluating
the data is to look at the highest Cell Average. This score would
typically be
located in the uppermost
RIGHT corner".
|
