Characterizing Learning As a Function of Attention Using a Multiple-Object Tracking Task: Defining Learning Trajectories in ASD and Other Neurodevelopmental Conditions

Background: Attention plays a significant role in learning by directing cognitive resources to satisfy task demands. As learning occurs, the need for attentional resources from the individual’s limited resource capacity decreases. Therefore, learning trajectories differ across the population as a function of differences in attentional capacities. These differences are exacerbated among individuals diagnosed with a neurodevelopmental condition (Melby-Lervåg & Hulme, 2015), whom present attention-specific difficulties (e.g., ADHD), or exhibit clinically significant difficulties in attention among other challenges (e.g., Autism Spectrum Disorder; ASD; Antshel, et al., 2013). Multiple Object-Tracking (MOT) paradigms are ideal for assessing the reciprocity between learning and attentional resource capacity because (i) MOT paradigms are robust measures of selective, sustained, distributed, and dynamic attention, and (ii) MOT can characterize individual differences in attention resource capacity (Tullo et al., 2018).

Objectives: We investigated whether a reciprocal nature of attention resource capacity and learning in ASD exists and if this relationship differs across other neurodevelopmental conditions that either do (i.e., ADHD) or do not have attentional difficulties (i.e., Intellectual or Learning Disability; ID; LD) as a primary concern. Specifically, we assessed the effect of repeated practice on an adaptive MOT paradigm, across 15 sessions. Additionally, we explored whether intelligence, our proxy for cognitive capability, influenced learning trajectories.

Methods: Children and adolescents (Ages 7-17; M_age=13.51; N=106) with confirmed diagnoses of either ASD (n=32), ADHD (n=35), or ID/LD (n=39) completed 15 daily MOT sessions. An MOT session involved visually tracking 3 of 8 spheres moving randomly for 8 seconds. Daily performance was defined as the average speed (cm/s) participants correctly tracked all target items. Cognitive status (or IQ) and attentional ability were assessed for all participants, using the Wechsler Abbreviated Scale of Intelligence – 2nd Edition (WASI-II) and the Conners Continuous Performance Test 3rd Edition (CPT-3), respectively.

Results: Collectively, the sample’s (i) IQ fell between 1 and 2 standard deviations below the population average (M_FSIQ=77.27, SD_FSIQ=13.16), and (ii) baseline attention met problematic levels of attention on the CPT-3 (M_d’t-score=60.00). A latent growth model revealed that MOT performance mapped onto a logarithmic function, which resembled a typical learning curve, at R²=0.87. Performance improved by 105% from the first to last day of testing. Moreover, the model revealed that day-one performance was predicted by intelligence: R²=0.28, and the rate of change, or learning trajectory differed across diagnostic groups. Here, the ASD group (M_sD1-15=1.11) demonstrated a greater standardized change compared to the ADHD (M_sD1-15=0.54) and ID/LD (M_sD1-15=0.52) groups.

Conclusions: These results characterize individual differences of learning capability and attention resource capacity, specific to children and adolescents diagnosed with a neurodevelopmental condition, where individuals with ASD demonstrated a unique learning trajectory. This unique trajectory may be indicative of distinct learning preferences specific to ASD cognitive profiles. Further, these findings demonstrate the value in using MOT to identify individual differences in learning trajectories as a function of attention. Therefore, we recommend (i) considering attention when characterizing learning capability on a case-by-case basis and (ii) adopting these descriptors of attention and learning towards tailoring learning material to capability.