Generalization of sensory auditory learning to top-down skills in a randomized controlled trial.

Abstract

Research has shown that auditory training improves auditory sensory skills; however, it is unclear whether this improvement is transferred to top-down skills, such as memory, attention, and language, and whether it depends on group characteristics in regard to memory and attention skills. The primary goal of this research was to investigate the generalization of learning from auditory sensory skills to top-down skills such as memory, attention, and language. We also aimed to compare whether this generalization process occurs in the same way among typically developing children and children with speech sound disorder. This study was a randomized controlled trial. Typically developing 7- to 12-yr-old children and children with speech sound disorder were separated into four groups: a trained control group (TDT; n = 10, age 9.6 ± 2.0 yr), a nontrained control group (TDNT; n = 11, age 8.2 ± 1.6 yr), a trained study group (SSDT; n = 10, age 7.7 ± 1.2 yr), and a nontrained study group (SSDNT; n = 8, age 8.6 ± 1.2 yr). Both trained groups underwent a computerized, nonverbal auditory training that focused on frequency discrimination, ordering, and backward-masking tasks. The training consisted of twelve 45 min sessions, once a week, for a total of 9 hr of training, approximately. Near-transfer (Gap-In-Noise [GIN] and Frequency Pattern Test) and far-transfer measures (auditory and visual sustained attention tests, phonological working memory and language tests) were applied before and after training. The results were analyzed using a 2 × 2 × 2 mixed-model analysis of variance with the group and training as the between-group variables and the period as the within-group variable. The significance threshold was p ≤ 0.05. There was a group × period × training interaction for GIN [F(1.35) = 7.18, p = 0.011], indicating a significant threshold reduction only for the TDT group (Tukey multiple comparisons). There was a significant group × period interaction [F(1.35) = 5.52, p = 0.025] and a training × period interaction for visual reaction time [F(1.35) = 4.20, p = 0.048], indicating improvement in the SSDT group and worsening in both nontrained groups. There was also a significant group × training × period interaction [F(1.35) = 4.27, p = 0.046] for the auditory false alarms, with a significant improvement after training only for the SSDT group. Analysis of variance also revealed that all groups exhibited approximately the same level of gains for all measures, except for GIN [F(3,38) = 4.261, p = 0.011] and visual response time [F(3.38) = 4.069, p = 0.014]. After training, the TDT group demonstrated a significant improvement for GIN and the SSDT exhibited the same for sustained attention, indicating learning generalization from an auditory sensory training to a top-down skill. For the other measures, all groups exhibited approximately the same level of gains, indicating the presence of a test-retest effect. Our findings also show that the memory span was not related to the learning generalization process given that the SSDT exhibited a more pronounced gain in attention skills after the sensory training.