Effects Of Training Sentence To Picture Matching

EFFECTS OF TRAINING SENTENCE TO PICTURE MATCHING AND OBJECT MANIPULATION TO IMPROVE SENTENCE COMPREHENSION IN APHASIA: ACQUISITION AND GENERALIZATION DATA ANALYSIS INTRODUCTION  In a previous study (Kiran, Caplan, et al., 2012), two treatments, one based on sentence to picture matching (SPM) and the other based on object manipulation (OM), that train patients on the relationship between syntactic structure and the meanings of sentences were developed.  Probe Baseline 1 Baseline 2 Baseline 3 RESULT 1: Which treatment is more effective (Tx Data) Object Manipulation Date of OM 3NP OC OR ORCNP PA UNACC Probe OM Version 40.00% 33.33% 0.00% 0.00% 6.67% 0.00% 12/8/2011 v1 50.00% 13.33% 0.00% 10.00% 20.00% 20.00% 12/13/2011 v2 30.00% 6.67% 20.00% 10.00% 20.00% 30.00% 12/20/2011 v3 Sentence-to-Picture Matching Date of 3NP OC OR ORCNP PA UNACC SPM Probe 80.00% 53.33% 26.67% 40.00% 66.67% 30.00% 12.1.2011 60.00% 40.00% 53.33% 50.00% 66.67% 20.00% 12/8/2011 80.00% 66.67% 46.67% 70.00% 66.67% 10.00% 12/13/2011 Probe 1 Probe 2 Probe 3 Probe 4 Probe 5 70.00% 50.00% 70.00% 70.00% 40.00% 40.00% 40.00% 40.00% 60.00% 60.00% 0.00% 13.33% 0.00% 0.00% 0.00% 10.00% 10.00% 0.00% 0.00% 0.00% 46.67% 66.67% 53.33% 66.67% 60.00% 20.00% 10.00% 10.00% 10.00% 0.00% 1/17/2012 1/31/2012 2/7/2012 2/21/2012 3/1/2012 v4 v1 v3 v2 v2 80.00% 70.00% 60.00% 50.00% 80.00% 46.67% 80.00% 53.33% 53.33% 33.33% 40.00% 46.67% 40.00% 40.00% 33.33% 60.00% 60.00% 40.00% 50.00% 40.00% 53.33% 53.33% 73.33% 46.67% 46.67% 30.00% 90.00% 70.00% 70.00% 80.00% Post Probe 1 Post Probe 2 Post Probe 3 60.00% 70.00% 50.00% 60.00% 60.00% 66.67% 0.00% 6.67% 6.67% 0.00% 0.00% 0.00% 73.33% 60.00% 66.67% 50.00% 30.00% 0.00% 3/6/2012 v4 3/8/2012 v1 3/13/2012 v3 80.00% 60.00% 70.00% 73.33% 80.00% 53.33% 40.00% 60.00% 46.67% 90.00% 50.00% 50.00% 66.67% 80.00% 66.67% 70.00% 50.00% 80.00% SPM Version Structure v2 structure 1 v3 structure 1 v4 structure 1  All patients improve as a function of treatment (p= 8.2E05) 1/11/2012 1/24/2012 1/31/2012 2/14/2012 2/28/2012 v1 SPM-unacc v2 SPM-unacc v3 SPM-unacc v4 SPM-unacc v1 SPM-unacc may need additional probe 3/1/2012 v4 structures 1 and 2 3/6/2012 v2 structures 1 and 2 3/8/2012 v3 structures 1 and 2  OM treatment more effective than SPM (NS for effect size, but significant for % change (p= 0.00917) We found the treatment to be effective in improving sentence comprehension of trained structures in fifteen patients with aphasia.  More patients improved on the OM task than SPM task. OBJECTIVES  In this study, we compare acquisition of trained structures and generalization to untrained structures and tasks across the two treatment approaches (SPM/OM).  Cross structure and cross task generalization is examined  We also examine effects of treatment on a broader array of sentences. RESULT 2: Generalization to untrained structures and tasks (Tx Data ) METHODS  N = 19 persons with aphasia  Participants were identified on the basis of two screening tests for syntactic comprehension (SPM & OM) with sentence structures ranging from object relative to active sentences.  A single subject multiple baseline design with order of task and structure counterbalanced across participants.  Sentence comprehension was trained on the affected sentence type in one task-related protocol generalization was examined to other structures. • Across structure within task generalization: Between OR <-> OC, OR <-> PASS, PASS <-> OC, OC -> UNACC. • Within structure across task generalization: From SPM OC -> OM OC RESULT 3: Generalization to untrained sentence types (Screener Task) Patients improved on their respective trained structures from pre-tx to post-tx screeners: (T=4.276, p<.001) RESEARCH POSTER PRESENTATION DESIGN © 2012 www.PosterPresentations.com 1. Greater changes from SPM to OM 2. Greater effects of treatment PASS to OC and OR Factor analysis on percent change: subcomponents reveal similarity in structure CONCLUSIONS • 19 patients underwent treatment. • 10 received OM treatment, 9 received SPM • Patients improve as a function of treatment although OM appeared to be more successful than SPM training. • Differences emerge between the two tasks (Salis & Edwards, 2009) • Across structure with task generalization • Between OR <-> OC, OR <-> PASS, PASS <-> OC, OC -> UNACC • Between different structures • Within structure across task generalization • No generalization from OM -> SPM • From SPM -> OM, for OC only. • Generalization to untrained structures on post-pre screener task • Results support the monitoring generalization effects • Training SPM results in greater cross task generalization than OM • Factor analysis reveals 9 components with eigenvalues > 1 and with significant correlations among components. • Of these Factors 1-5 explain 65% of the variance • Factors reveal structures with similar structure and movement tend to change together as a function of treatment