Signed Paired Associates Test (SPAT) SPAT Structure

Signed Paired Associates Test (SPAT)

SPAT Structure n Similar to WMS “paired associates” subtest n 14 sign pairs – 7 easy & 7 hard n Based on sign associate frequency research n Immediate recall phase (4 learning trails) n Delayed recall phase (free, then cued) n 9 primary scores n 3 immediate recall n 6 delayed recall

SPAT Studies n De. Matteo, Pollard, & Lentz, 1987 n Initial norms, negative correlation with age n Pollard, Rediess, & De. Matteo, 2005 n 38 healthy deaf adults n n n Mean age 27. 7 (s. d. , 4. 8, range 18 -34) 55% male, 45% female 35 deaf adults referred for neuropsych. testing n n Mean age 30. 6 (s. d. , 8. 9, range 18 -57) 59% male, 41% female

Pollard, Rediess & De. Matteo, 2005 n Healthy sample n Screened for neurological deficits n WAIS-R PIQ (required >70 to participate) n SPAT, ASL Stories Test administered n Clinical sample n Suspected of brain impairment n PIQ or Ravens IQ > 70 required for study n SPAT and other tests deemed necessary

2005 SPAT Study Results n Age of two samples not significantly different n Mean IQ differed (p =. 007) n Healthy 103. 9 (s. d. , 13. 0, range 75 -128) n Clinical 94. 3 (s. d. , 16. 1, range 70 -124 n Performance on nine SPAT scores very similar to De. Matteo, Pollard, & Lentz, 1987

SPAT Norms Immediate Recall Delayed Free + Cued Recall Easy Total Hard Total Sum Total 28 28 56 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 De. Matteo, et al.

2005 SPAT Results (cont. ) n All 13 scores (9 primary scores and 4 trial-by- trail learning totals) significantly differed between healthy and clinical groups. n Learning curves evidenced for both groups but harder for clinical sample n PIQ positively correlated with all 9 primary SPAT scores

Healthy v. Clinical Performance

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. ) n Retention scores n Recall (free and delayed) expressed as percentage of total learned by trial 4 n 69% retention at delayed free recall n ~100% retention delayed fee + cued recall n These percentages the same for both groups n No significant differences in retention scores

2005 SPAT Results (cont. ) n Forward step-wise discriminate analysis n What contributed most to SPAT performance? n 7 of 9 primary SPAT scores & PIQ/Ravens n Final analysis included: n n 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” n Performance patters similar to WMS P. A. and other “hearing” verbal tests Improved retention over learning trials n Semantically related easier than non-related n PIQ positively correlated with performance n Age negatively correlated with performance (De. Matteo, et al. , 1987 and pilot study only) n n These findings speak to construct validity

Construct and Discriminate Validity n In every performance indicator tested, the clinical sample performed more poorly than the healthy sample n 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 n Sensitive but not too specific = more useful test

Future Research & Clinical Ideas n Norms needed for elderly and children! n Interpreted vs. direct administration n Correlation with education n Other clinical samples n Deaf subpopulations (e. g. at risk etiologies) n Performance of those with less ASL fluency n Correlation with non-verbal learning tests n Correlation with “hearing” verbal learning tests n Altered administration (voice, length, delay period)