WEBVTT

00:00:05.276 --> 00:00:06.076
I am Don Compton.

00:00:06.226 --> 00:00:08.366
I am an associate professor
of special education

00:00:08.416 --> 00:00:09.686
at Vanderbilt University.

00:00:10.136 --> 00:00:12.716
I am sure districts and schools
are wondering what it takes

00:00:12.716 --> 00:00:14.646
to implement an effective
screening,

00:00:14.836 --> 00:00:16.315
a universal screening procedure.

00:00:16.315 --> 00:00:18.816
And what we recommend is

00:00:18.816 --> 00:00:23.666
that first they develop a
schoolwide team and they come

00:00:23.666 --> 00:00:25.096
up with a strategic plan,

00:00:25.096 --> 00:00:27.726
and then that plan
maybe is multi-year

00:00:27.726 --> 00:00:30.406
in terms of rolling this out.

00:00:30.406 --> 00:00:34.896
Members on that team could
include classroom teachers,

00:00:34.976 --> 00:00:39.216
an administrator, psychologist,
special ed teachers,

00:00:39.336 --> 00:00:41.736
paraprofessionals,
all sorts of people.

00:00:41.896 --> 00:00:44.536
They're going to need people
who have expertise in screening

00:00:44.676 --> 00:00:50.186
and in testing, those who can
actually go out and assess kids.

00:00:50.186 --> 00:00:52.466
Universal screening
lasts for-depending

00:00:52.466 --> 00:00:53.476
on how big your school is,

00:00:53.476 --> 00:00:56.456
how many kids-it may take a
few days, it may take a week.

00:00:56.856 --> 00:00:56.866


00:00:58.686 --> 00:01:01.386
Once all kids have been
screened, then the process needs

00:01:01.536 --> 00:01:06.906
to go forward with putting data
into the system, making decision

00:01:06.906 --> 00:01:10.746
on cut scores, looking at that,
making sure it makes sense,

00:01:10.786 --> 00:01:12.356
bringing teacher
judgment to make sure

00:01:12.356 --> 00:01:15.896
that the kids we've identified
fit within what they think

00:01:15.896 --> 00:01:17.846
about is going on
in their classroom.

00:01:17.846 --> 00:01:21.976
And the system is
going to take money.

00:01:21.976 --> 00:01:22.266


00:01:24.346 --> 00:01:26.616
Let me give you an example
of how a school might set

00:01:26.686 --> 00:01:28.466
up a universal screening system.

00:01:28.516 --> 00:01:29.616
I think it would be good

00:01:29.616 --> 00:01:31.326
to start the year
before by planning.

00:01:31.436 --> 00:01:34.246
The school needs to
identify what type

00:01:34.246 --> 00:01:37.306
of universal screener they want
to administer, who is going

00:01:37.306 --> 00:01:39.826
to administer that measure,
who is going to score,

00:01:39.956 --> 00:01:41.136
who is going to set cut-points.

00:01:41.286 --> 00:01:43.736
The first year of
universal screening,

00:01:43.736 --> 00:01:48.486
probably what schools can do
best is just give the screener

00:01:48.566 --> 00:01:49.416
to all children.

00:01:49.666 --> 00:01:54.736
They're going have to rely on
some kind of cut-point benchmark

00:01:54.766 --> 00:01:58.676
that is set by the
screening vendor.

00:01:58.676 --> 00:02:02.806
And that might be the best
they can do the first year.

00:02:02.866 --> 00:02:06.516
We recommend then that
schools slowly adopt a system

00:02:06.616 --> 00:02:10.545
where they give the universal
screener, identify the kids

00:02:10.616 --> 00:02:13.986
who are at risk-fairly
liberally-and then progress

00:02:13.986 --> 00:02:17.066
monitor over time to
really figure out the kids

00:02:17.066 --> 00:02:19.056
who are not benefiting
from classroom instruction

00:02:19.056 --> 00:02:20.236
and are truly at risk.

00:02:20.386 --> 00:02:22.106
So that might be the
second or third year.

00:02:22.676 --> 00:02:25.236
Then a school might
decide, okay,

00:02:25.236 --> 00:02:27.556
after we've implemented
universal screening,

00:02:27.596 --> 00:02:30.476
progress monitoring, we're
going to watch and follow

00:02:30.526 --> 00:02:32.526
up the outcomes of kids

00:02:32.616 --> 00:02:34.536
to see how good our
screener is working.

00:02:34.856 --> 00:02:37.196
That gives you better
data as to who you missed,

00:02:37.316 --> 00:02:39.356
how many extra kids
you are identifying.

00:02:39.706 --> 00:02:44.266
And then maybe in the last
phase, they would use that data

00:02:44.266 --> 00:02:46.116
to actually refine
their own cut-points,

00:02:46.116 --> 00:02:48.716
set their own district
or school cut-points,

00:02:48.716 --> 00:02:52.186
and then they have a much
better screener that's tailored

00:02:52.266 --> 00:02:52.976
to their population.

00:02:54.586 --> 00:02:54.976


00:02:56.046 --> 00:02:58.096
Some may be wondering
what a cut score is.

00:02:58.096 --> 00:03:01.946
A cut score can be used either
in a benchmarking system

00:03:02.016 --> 00:03:03.856
or in a progress
monitoring system.

00:03:04.066 --> 00:03:07.996
It's just a dichotomized
score, either yes or no,

00:03:07.996 --> 00:03:10.806
of whether kids are
considered to be at risk or not.

00:03:10.996 --> 00:03:15.076
So you give your universal
screener or you give a series

00:03:15.076 --> 00:03:18.336
of progress monitoring probes,
and you score the children

00:03:18.376 --> 00:03:21.156
on that and then you just
decide, does the child score

00:03:21.156 --> 00:03:23.496
above or below a
critical cut score

00:03:23.776 --> 00:03:25.186
that you, the school, sets.

00:03:25.396 --> 00:03:28.166
So our objective is
to try to identify

00:03:28.236 --> 00:03:31.346
with 100 percent accuracy those
children who are going

00:03:31.346 --> 00:03:33.706
to have future reading
problems and those who aren't.

00:03:33.966 --> 00:03:35.576
That way, we can
give the children

00:03:35.696 --> 00:03:38.306
who are having problems
structured intervention

00:03:38.336 --> 00:03:41.056
to help their reading outcomes.

00:03:41.056 --> 00:03:42.666
So what does that mean?

00:03:42.866 --> 00:03:43.736
How does that work?

00:03:43.736 --> 00:03:47.356
Well, we're trying to predict
behavior in the future,

00:03:47.486 --> 00:03:48.966
which is not an easy
thing to do.

00:03:50.586 --> 00:03:50.976


00:03:52.056 --> 00:03:54.156
In terms of trying to
pick the right children,

00:03:54.156 --> 00:03:56.646
it's impossible to
be 100 percent accurate.

00:03:56.886 --> 00:04:01.136
So we have this cut score that
we can use to put kids into risk

00:04:01.286 --> 00:04:03.196
and not-at-risk classes.

00:04:03.196 --> 00:04:04.926
That cut-point is
fairly arbitrary.

00:04:04.926 --> 00:04:07.716
Schools can move
that up or down.

00:04:07.956 --> 00:04:10.886
A more lenient cut score
means you have a better chance

00:04:10.886 --> 00:04:13.266
of getting all the children
who are going to have problems.

00:04:13.466 --> 00:04:15.756
However, what happens
is you get a lot

00:04:15.756 --> 00:04:17.326
of what are called
false positives.

00:04:17.326 --> 00:04:20.815
Those are kids who score below
the cut-point, making them seen

00:04:20.815 --> 00:04:24.036
as if they are at risk, but they
develop normal reading skills.

00:04:24.036 --> 00:04:26.716
So you would be putting children
who don't need intervention

00:04:27.186 --> 00:04:28.666
into structured intervention,

00:04:28.666 --> 00:04:30.566
which costs money
and is inefficient.

00:04:30.966 --> 00:04:33.886
You can make the
cut-point more stringent,

00:04:33.886 --> 00:04:37.516
and what happens then is
you miss some of your kids

00:04:37.576 --> 00:04:40.696
but you put less kids into
tutoring, but the ramification

00:04:40.746 --> 00:04:43.616
of that is that kids who need
tutoring don't get tutoring.

00:04:43.816 --> 00:04:49.186
Deciding on a proper benchmark
for a screen or cut-point

00:04:49.286 --> 00:04:51.176
for a screen or progress
monitoring is a

00:04:51.286 --> 00:04:52.516
difficult process.

00:04:52.626 --> 00:04:54.776
It's one that schools
are going to have

00:04:54.776 --> 00:04:56.346
to struggle with a little bit.

00:04:56.516 --> 00:05:00.576
Initially, I think most
vendors will have cut-points

00:05:00.686 --> 00:05:03.216
or benchmarks for their measures

00:05:03.326 --> 00:05:07.356
that schools can
immediately use.

00:05:07.526 --> 00:05:10.596
The issue with these is you're
not sure whether those are going

00:05:10.596 --> 00:05:12.766
to generalize to your
school population.

00:05:13.116 --> 00:05:15.836
So we recommend that
schools, if possible,

00:05:16.076 --> 00:05:17.866
start to develop
their own cut-points

00:05:17.996 --> 00:05:19.686
for their populations.

00:05:19.736 --> 00:05:21.716
This may take time.

00:05:21.896 --> 00:05:26.166
They may have to start with the
vendor's suggested cut-points

00:05:26.166 --> 00:05:27.836
and modify those over time.

00:05:28.036 --> 00:05:30.796
Schools have to kind
of play the balance off

00:05:30.796 --> 00:05:33.206
of how many kids can they
afford to give tutoring

00:05:33.296 --> 00:05:35.596
and how many kids are
they willing to miss.

00:05:35.596 --> 00:05:40.546
And in general, we try to-all
things being equal-we try

00:05:40.956 --> 00:05:43.676
to identify as many
of the true positives

00:05:43.676 --> 00:05:46.596
as we can, and as we can afford.

00:05:46.596 --> 00:05:48.426
And we'll bring some
kids into tutoring

00:05:48.506 --> 00:05:49.926
that maybe don't need it,

00:05:50.086 --> 00:05:51.826
but usually it's a good
experience for them.