Signed Paired Associates Test (SPAT) SPAT Structure Similar to WMS “paired associates” subtest 14 sign pairs – 7 easy & 7
Download
Report
Transcript Signed Paired Associates Test (SPAT) SPAT Structure Similar to WMS “paired associates” subtest 14 sign pairs – 7 easy & 7
Signed Paired Associates Test
(SPAT)
SPAT Structure
Similar to WMS “paired associates” subtest
14 sign pairs – 7 easy & 7 hard
Based on sign associate frequency research
Immediate recall phase (4 learning trails)
Delayed recall phase (free, then cued)
9 primary scores
3 immediate recall
6 delayed recall
SPAT Studies
DeMatteo, Pollard, & Lentz, 1987
Initial norms, negative correlation with age
Pollard, Rediess, & DeMatteo, 2005
38 healthy deaf adults
Mean age 27.7 (s.d., 4.8, range 18-34)
55% male, 45% female
35 deaf adults referred for neuropsych. testing
Mean age 30.6 (s.d., 8.9, range 18-57)
59% male, 41% female
Pollard, Rediess & DeMatteo, 2005
Healthy sample
Screened for neurological deficits
WAIS-R PIQ (required >70 to participate)
SPAT, ASL Stories Test administered
Clinical sample
Suspected of brain impairment
PIQ or Ravens IQ > 70 required for study
SPAT and other tests deemed necessary
2005 SPAT Study Results
Age of two samples not significantly different
Mean IQ differed (p = .007)
Healthy 103.9 (s.d., 13.0, range 75-128)
Clinical 94.3 (s.d., 16.1, range 70-124
Performance on nine SPAT scores very similar
to DeMatteo, Pollard, & Lentz, 1987
SPAT Norms
Immediate Recall
Delayed Free Recall
Delayed Free + Cued
Recall
Easy
Total
Hard
Total
Sum
Total
Easy
Total
Hard
Total
Sum
Total
Easy
Total
Hard
Total
Sum
Total
28
28
56
7
7
14
7
7
14
26.5
(2.4)
18.4
(5.7)
44.9
(7.4)
4.6
(1.2)
4.6
(1.7)
9.2
(2.5)
6.9
(0.4)
5.8
(1.7)
12.7
(1.9)
25.7
(4.0)
18.0
(6.2)
43.7
(9.4)
4.5
(1.2)
4.4
(1.9)
8.9
(2.8)
6.8
(0.9)
5.7
(2.1)
12.5
(2.8)
Maximum Possible
Current study
DeMatteo, et al.
2005 SPAT Results (cont.)
All 13 scores (9 primary scores and 4 trial-by-
trail learning totals) significantly differed
between healthy and clinical groups.
Learning curves evidenced for both groups
but harder for clinical sample
PIQ positively correlated with all 9 primary
SPAT scores
**
Tr
ia
l1
**
T
*T
r ia otal
(1
** l 2
To 4)
*T
tal
r ia
l
(1
**
3
To 4)
*T
** r ial tal (
Im
1
4
To 4)
m
.
tal
**
*I Eas
(1
y
m
To 4)
m
t
** . H
ar al (2
**
*I
d
m
De
To 8)
m
la
.S
ta
um l ( 2
** yed
De
Fr
To 8)
ee
la
tal
Ea
** yed
(5
sy
Fr
6)
*D Del
T
ee
o
a
ela
ye
tal
Ha
d
y
(7
rd
ed
Fr
**
)
e
T
De
Fr
e
o
Su
tal
ee
la
+
y
m
(7
**
e
Cu
T
)
De d F
ed
ot
r
al
ee
la
ye
+C Eas
(1
4)
y
d
u
To
Fr
ed
ee
t
+C Har al (
7
d
ue
To )
d
ta
Su
l(
m
To 7)
ta
l(
14
)
Mean Pairs Recalled
Healthy v. Clinical Performance
49
42
35
28
Healthy
21
Clinical
14
7
0
SPAT-PIQ Correlations
SPAT Score
r value
p value
Immediate Recall Easy Total
.30
.011
Immediate Recall Hard Total
.47
<.001
Immediate Recall Sum Total
.45
<.001
Delayed Free Recall Easy Total
.48
<.001
Delayed Free Recall Hard Total
.46
<.001
Delayed Free Recall Sum Total
.50
<.001
Delayed Free + Cued Recall Easy Total
.27
.021
Delayed Free + Cued Recall Hard Total
.49
<.001
Delayed Free + Cued Recall Sum Total
.46
<.001
2005 SPAT Results (cont.)
Retention scores
Recall (free and delayed) expressed as
percentage of total learned by trial 4
69% retention at delayed free recall
~100% retention delayed fee + cued recall
These percentages the same for both groups
No significant differences in retention scores
2005 SPAT Results (cont.)
Forward step-wise discriminate analysis
What contributed most to SPAT performance?
7 of 9 primary SPAT scores & PIQ/Ravens
Final analysis included:
Immediate recall hard total
PIQ/Ravens IQ
Delayed free + cued recall hard total
Consistent finding that learning and retention
of hard pairs is most clinically salient aspect
“It acts like we expect a verbal
learning and memory test to act”
Performance patters similar to WMS P.A. and
other “hearing” verbal tests
Improved retention over learning trials
Semantically related easier than non-related
PIQ positively correlated with performance
Age negatively correlated with performance
(DeMatteo, et al., 1987 and pilot study only)
These findings speak to construct validity
Construct and Discriminate Validity
In every performance indicator tested, the
clinical sample performed more poorly than the
healthy sample
Finding that immediate and delayed recall total
hard scores best differentiated the two samples
parallels research showing that semantically
unrelated word pair learning is a sensitive
measure of memory impairment in hearing
clinical samples and healthy elderly people
Sensitive but not too specific = more useful test
Future Research & Clinical Ideas
Norms needed for elderly and children!
Interpreted vs. direct administration
Correlation with education
Other clinical samples
Deaf subpopulations (e.g. at risk etiologies)
Performance of those with less ASL fluency
Correlation with non-verbal learning tests
Correlation with “hearing” verbal learning tests
Altered administration (voice, length, delay period)