Temporal processing tele-intervention improves language,attention,and memory in children with neurodevelopmental disorders

Abstract

Objective

Temporal processing is the brain's ability to process rapid successive stimuli, and children with neurodevelopmental disorders showed temporal processing deficits. Empirical evidence suggests that in-person intervention on temporal processing improves various cognitive functions of these children, and the present study aimed to study the effects of temporal processing tele-intervention (TPT) on the cognitive functions of children with neurodevelopmental disorders.

Methods

Ninety-five children with neurodevelopmental disorders were recruited and randomly assigned to remotely receive either TPT or conventional language remediation (CLR) in 20 parallel group-based intervention sessions once per week. Their cognitive functions were assessed before and after the intervention.

Results

The TPT group demonstrated a specific and significant improvement in working memory (p < .001). While there was an overall significant improvement in sustained attention in terms of processing speed after both types of intervention (p = .006), the positive effects of TPT might be more prominent than that of CLR given the significant pre–post difference after receiving TPT (p = .012) but not CLR (p = .21). Regarding rapid naming accuracy which had marginally significant improvement after the intervention (p = .03), the trend of improvement in TPT (p = .05) also seemed more apparent than that of CLR (p = .18). Finally, the TPT group had significant improvement in word knowledge (p < .001), rapid naming speed (p < .001), sustained attention in terms of accuracy (p < .001), and verbal learning and memory (p < .01) to an extent similar to that of the CLR group.

Conclusions

These findings suggest that TPT can be a potential intervention for improving cognitive functions in children with neurodevelopmental disorders.

Clinical trial registration number: NCT05428657 at ClinicalTrials.gov (https://clinicaltrials.gov/).

Keywords

Tele-intervention temporal processing cognitive improvement children neurodevelopmental disorders

Introduction

Temporal processing refers to an individual's time resolution when processing stimuli presented in rapid succession on a timescale as short as milliseconds. It is an essential perceptual processing skill fundamental to many cognitive functions, such as perception, attention, memory, language, and motor abilities.^1–4 Temporal processing deficits are common in children with neurodevelopmental disorders. As early as the 1970s, Tallal and Piercy⁵ reported that aphasic children demonstrated a deficiency in distinguishing and judging the temporal order of tone pairs. Since then, an increasing number of studies have further demonstrated temporal processing deficits in other neurodevelopmental disorders, including autism spectrum disorder (ASD)^6–9 and attention-deficit/hyperactivity disorder (ADHD).^10–13 Though inconsistent findings of relatively intact temporal processing for nonspeech stimuli in ASD^14,15 and ADHD^16,17 have been reported, many studies have put forward that temporal processing deficits found in ASD and ADHD may be a possible factor mediating their cognitive impairment in attention, episodic memory, and language.^18–22 For instance, Maister and Plaisted-Grant¹⁹ found that when children with ASD were required to reproduce stimuli with varying durations (a common paradigm testing temporal processing), they committed significantly more errors than typically developed children, and their temporal processing deficit might be the underlying cause of impaired episodic memory. Foss-Feig et al.²⁰ also reported an association between impairment in rapid auditory temporal processing and difficulties in language processing among children with ASD in a gap detection task. In addition, for individuals with ADHD, Hart et al.²¹ revealed that temporal processing tasks were sensitive in classifying children with ADHD from their normal counterparts, and the temporal processing measure was significantly correlated with the core inattention and hyperactive symptoms of ADHD. Temporal processing was also intercorrelated with impaired executive functions in ADHD, such as working memory and attention. However, timing deficits in ADHD remained to exist even when executive functions were controlled, suggesting independent deficits in the temporal domain.²²

To evaluate temporal processing, one of the well-known experimental paradigms is the temporal order judgment (TOJ) paradigm,²³ in which children with neurodevelopmental disorders usually require longer interstimulus intervals (ISIs) to separately perceive two acoustic stimuli that are presented shortly one after the other^5,6,8,11,12 and demonstrate lower accuracy in sequencing the two stimuli.^5,6,10,12 In specific, children with ASD required 48% longer ISIs for reliable repetition of the sound sequences than typically developed children.⁸ Children with ADHD showed a higher TOJ threshold for tones¹¹ and lower accuracy in sequencing tonal and consonant–vowel (CV) syllabic pairs.¹⁰ Two recent studies applying an online temporal processing assessment revealed similar temporal processing deficits in children with ADHD¹² and ASD,⁶ suggesting that children with ADHD or ASD showed significantly lower accuracy rates than their typically developed counterparts in TOJ tasks using tonal and CV syllabic pairs and a higher ISI passing threshold for CV pairs.

Given the deficits in temporal processing commonly found in children with neurodevelopmental disorders, many studies have explored the effectiveness of temporal processing intervention (TPI) in improving their cognitive functions. Empirical evidence suggests that TPI has a positive effect on language abilities. By implementing auditory TPI for 6 weeks,²⁴ Tallal et al. demonstrated positive outcomes of TPI on speech discrimination and comprehension in children with language-learning impairment. Improvement has also been shown in phonological processing,^25,26 rapid naming,²⁷ spelling,²⁸ and reading ability²⁸ in children with dyslexia after TPI. Specifically, Wang et al.²⁷ reported that Chinese children with dyslexia took a shorter time to complete a rapid digit naming task after receiving 12 sessions of TPI for 3 to 4 weeks. Strehlow et al.²⁸ also found a 4-week TPI to have positive effects on improving reading and spelling in children with dyslexia.

Apart from language, improvements in other cognitive domains, including attention and memory, were shown after TPI. For instance, Stevens et al.²⁹ found an improvement in selective auditory attention after 6 weeks of TPI in children with language impairment. In adults, Szelag and Skolimowska³⁰ reported that TPI improved the sustained and divided attention, visual attention, and memory functions of healthy older adults. In addition, an information processing training program with TOJ-based training of rapidly successive sweeps as one of the six training exercises enhanced global auditory memory scores.^31,32

While the majority of studies on TPI are targeted at children with language impairment or dyslexia and older adults,^24–32 the effect of TPI on children with ASD and ADHD remains largely unknown. ASD and ADHD are commonly found to have multiple cognitive impairments, and temporal processing is a fundamental skill mediating many cognitive functions. Based on the above empirical evidence for the presence of temporal processing deficits in ASD^6–9 and ADHD^10–13 and the possible association between their temporal processing deficits and various cognitive impairments,^18–22 it is therefore postulated that training on temporal processing may also enhance the language, attention, and memory functions of children with ASD and ADHD.

Besides, previous studies on TPI were primarily conducted through in-person intervention; therefore, the effectiveness of temporal processing tele-intervention (TPT) is still being determined. With the vigorous development of tele-intervention in recent years, some studies have shown that even children with neurodevelopmental disorders could benefit from remote online intervention.^33–35 Tele-intervention that can be delivered remotely makes intervention more easily accessible for children with neurodevelopmental disorders in lower-income classes or remote areas. As a result, more children with neurodevelopmental disorders can benefit from tele-intervention. Thus, the present study aims to examine the potential effects of TPT on children with ASD and ADHD. With the positive outcomes of TPI on children and adults, it is anticipated that TPT would show similar benefits as TPI.

The TPT program adopted in the present study was developed by the first author. It is a web-based computerized intervention program that can be administered remotely on a computer or tablet device. The intervention program was designed based on the TOJ task of identifying and ordering pairs of rapidly successive tone and CV stimuli. Since 2019, our team has applied the TPT program to train over 100 children with neurodevelopmental disorders in Hong Kong, and positive outcomes were observed in terms of more accurate temporal processing during the intervention, better verbal expression and comprehension, reading and spelling skills, and academic performance in real life.

Given our encouraging clinical observations and the empirical support for the potential effects of TPI, the present study aims to examine whether web-based TPT can enhance the language, attention, and memory functions of children with neurodevelopmental disorders. Based on previous empirical evidence for an association between temporal processing and cognitive functions^1–4 and its deficit associated with cognitive impairments in ASD and ADHD,^18–22 it is hypothesized that training the fundamental temporal processing skills could enhance various cognitive functions of children with neurodevelopmental disorders. Besides, the effects of TPT on cognitive functions will be compared with those of conventional language remediation (CLR). CLR was chosen as the comparison group because it is one of the remedial supports commonly provided for children with neurodevelopmental disorders in their formal schools in Hong Kong. By comparing the training effects of CLR, the present study may shed some light on the possibility of applying TPT as an alternative intervention for children with neurodevelopmental disorders.

Methods

Participants

In total, 95 children with neurodevelopmental disorders aged 6 to 11 years in grades 2 to 6 were recruited and referred by their school teachers from 4 local mainstream primary schools. The sample size was predetermined based on the mean effect size of 0.82 for applying TPI to enhance the cognitive functions of children with language impairment, as outlined in Tallal's study.³⁶ Power analysis was conducted to compute the total sample size per intervention group to be at least 31, with the power at 0.95 and alpha value at .05. Accounting for an approximate 35% attrition rate, a sample size of 48 children per group was required. According to the Education Bureau records, these children were formally diagnosed with ASD, ADHD, specific learning difficulties (SpLD), and/or a borderline level of intellectual functioning. The diagnosis was formulated by a psychiatrist, a pediatrician, a clinical psychologist at a general hospital or child assessment center, or an educational psychologist at school. None of them reported a history of head injury or epilepsy. All children participated voluntarily, with written informed consent obtained from their parents before the study. This study was conducted in compliance with the requirements stated in the Declaration of Helsinki of the World Medical Association. The research protocol was approved by the Joint Chinese University of Hong Kong–New Territories East Cluster Clinical Research Ethics Committee (CREC reference number: 2020.501-T).

Procedure

At baseline, children were assessed individually on their cognitive functions, including language, working memory, sustained attention, and verbal learning and memory, using standardized neuropsychological tests and an experimental paradigm by trained research assistants at their primary schools.

After baseline assessment, they were randomly assigned to receive either TPT (n = 47) or CLR (n = 48) for 20 intervention sessions, 75 minutes each, once per week, over 8 months, which was the usual dosage and duration of remedial support for children with neurodevelopmental disorders in Hong Kong. Randomization was performed by a research assistant using a random number generator. The two intervention groups were matched based on age (t(93) = −0.68, p = .50), gender (χ²(1) = 0.04, p = .85), years of education (t(93) = −0.38, p = .70), and diagnosis (χ²(3) = 4.29, p = .23). Each intervention program was delivered to a group of 10 to 15 children remotely through a video conferencing platform. Each group of intervention was run by a team of one trained instructor with two to three assistants at the same time in a parallel session in separate virtual meeting rooms. The children attended the online intervention sessions at their home or school. In each meeting room, children of the same group met each other virtually. During the training, the instructor was responsible for providing tele-intervention and working with the assistants to help children who had encountered technical problems or difficulties in accomplishing the training tasks.

After the intervention, the children were reassessed with the same tests given at baseline to evaluate the effects of TPT and CLR on their cognitive functions. The assessments were conducted by trained research assistants who were blinded to group assignments. A total of 20 students were excluded due to dropout (n = 6), missing or invalid data (n = 4), or an attendance rate below 25% (n = 10). In addition, the majority of children in the TPT (95%) and CLR (87%) groups were diagnosed with ASD or ADHD. Yet, only a limited number of children with SpLD (n = 5 in the CLR group and n = 1 in the TPT group) and borderline IQ (n = 1 in the TPT group) eventually completed the intervention. They were not equally distributed in the two intervention groups. To reduce heterogeneity between the two intervention groups, seven children with SpLD and borderline IQ were excluded. Finally, 35 children remained in the TPT group and 33 children in the CLR group for data analysis.

Outcome measures

Language

Rapid Automatic Naming (RAN) is an experimental paradigm adopted from Ebaid and Crewther's study that measures rapid letter naming.³⁷ Participants were asked to read aloud a 6 × 5 array of letters as fast and accurately as possible in 60 s. The total correct score and total errors were computed as separate measures of rapid naming speed and accuracy, respectively. Furthermore, the Chinese Vocabulary Test (CVT) was used to measure Chinese word knowledge.³⁸ Participants were required to verbally explain the meaning of 18 Chinese words. Each item was scored as “0”, “1”, or “2” and then added together to form a total score.

Working memory

The Number Span Test (NST) with forward (NST-F) and backward (NST-B) conditions was adopted to assess working memory.³⁹ Participants were read sequences of digits with an increasing span (from 2 to 12) and had to repeat in the same order as presented in the forward condition or reverse order in the backward condition. NST-B requires mental effort to manipulate information, whereas NST-F simply requires holding information temporarily without manipulation. Each condition had maximally 22 trials, and every trial with a correct answer earned one mark. The total score for each condition was computed separately by summing up the number of correct trials.

Sustained attention

The Digit Cancellation Test (DCT) was adopted to measure sustained attention requiring rapid visual tracking and accurate selection of target stimuli.³⁹ Participants were asked to cancel out as many “3” digits as possible in a two-page array of randomized digits (0–9) in 10 minutes. The total time (in seconds) taken to complete the task and the total score (i.e. correct cancelation minus commission errors) were computed as separate measures of processing speed and accuracy, respectively.

Learning and memory

The Hong Kong List Learning Test (HKLLT) was adopted to measure verbal episodic memory.⁴⁰ Participants were required to learn a verbally presented list of 16 two-character Chinese words across 3 learning trials, and then recall the words from memory after 30 minutes. The total correct recall across learning trials and the correct delayed recall were computed as separate measures of verbal learning and memory, respectively.

Interventions

Temporal processing tele-intervention

The online TPT program, developed by our research team, is operated by the Pro-talent Association Ltd in Hong Kong. It is a tele-intervention program that can be run remotely using a computer or tablet device. The tele-intervention program comprises two modules of TOJ-based intervention using tone or CV syllabic pairs separated by varying ISIs (ranging from 500 to 0 ms). In the tone training, each trial presents one of the four possible tone pairs (U-U, U-D, D-D, and D-U) composed of an upward-gliding (U) and/or a downward-gliding (D) tone. In the CV training, there are 36 possible CV syllabic pairs formed from 6 stop-consonant syllables (/ba/, /da/, /ga/, /ka/, /pa/, and /ta/) that involve a rapid formant transition (i.e. rapid frequency changes in the initial tens of milliseconds of the syllable). These syllables were adopted from the standard Bergen dichotic listening paradigm.⁴¹

Each tele-intervention module begins with a practice session to familiarize the trainee with the sound stimuli and the intervention requirements. After passing the practice session, the program automatically proceeds to the training session. The training progresses in difficulty in a three-down-one-up adaptive procedure as the trainee's performance improves. That is, after every three consecutive correct responses, the trainee earns a star and is upgraded to a higher training sublevel with a reducing ISI from 500 to 0 ms; on the other hand, every single error would increase the ISI by one step. The tone and CV stimuli variants are synthesized using Audacity (version 2.2.2)⁴² and Praat (version 6.1.11)⁴³ software. Completing every 20 steps of ISI variants leads to a shortened tone duration in the tone training and a reduced extension and amplification of formant transitions in the CV training. The ultimate goal is to accurately discriminate and sequence the tone at the shortest ISI and tone duration in the tone training and the initial CV syllables without modification in the CV training.

Throughout the training, the participants could replay the stimuli presentation anytime and received guidance from the instructor and assistants. They could choose to continue or quit the training program at any time. Responses were scored automatically, and feedback was provided for each trial. In the present study, the progress of intervention was measured in terms of the percentage of completion (maximally 100%) achieved by the participants. A higher percentage of completion reflects a better ability in temporal processing of auditory stimuli presented in rapid succession.

Conventional language remediation

CLR is composed of interactive teaching and computerized exercises focused on training Chinese reading and writing skills. The training materials were adopted from three standard teaching/training resources published by the Hong Kong Specific Learning Difficulties Research Team,⁴⁴ the Department of Special Education and Counselling at the Hong Kong Institute of Education,⁴⁵ and the Education Bureau.⁴⁶ They involved game-like computerized training programs to teach word recognition,⁴⁴ an online animated training program to teach narrative skills and reading comprehension,⁴⁶ and some mini-games and paper-and-pencil exercises to build Chinese vocabulary and writing skills.⁴⁵

Data analysis

The baseline characteristics and cognitive functions were compared between intervention and diagnostic groups using independent sample t-tests or chi-squared tests. When a confounding variable was identified, analysis of covariance (ANCOVA) was performed to control its effect on the dependent variable. The effects of TPT and CLR on each cognitive domain were compared across time by separate two-way mixed analysis of variance (ANOVA) with time (pre vs. post) as a within-subject factor and Group (TPT vs. CLR) as a between-subject factor. Further post hoc pairwise comparisons using paired t-tests were conducted to examine the specific changes in each group. Bonferroni-adjusted p-value at .025 for family-wise errors was applied to measures from the same cognitive test in mixed ANOVA and paired t-test.

Moreover, their pre–post difference scores (i.e. postmeasurement minus premeasurement) were compared between groups using independent t-tests. Partial eta squared (η_p²) and Cohen's d were presented to indicate the effect sizes. All statistical analyses were performed using SPSS 28.0 software.

Results

Baseline characteristics and cognitive performance

Table 1 presents the baseline demographic and clinical characteristics and level of cognitive functions of the two intervention groups. Both groups were matched based on age (t(66) = 0.12, p = .90), education level (t(66) = 0.33, p = .74), gender (χ²(1) = 0.19, p = .66), and diagnosis (χ²(1) = 0.30, p = .58). They also had comparable baseline cognitive functions (p > .05) (Table 1).

Table 1.

Participants’ characteristics and cognitive performance at baseline.

Baseline characteristics/cognitive performance	CLR (n = 33)		TPT (n = 35)		t/χ²	p
Baseline characteristics/cognitive performance	M	SD	M	SD	t/χ²	p
Age (years)	8.44	1.00	8.47	1.13	0.12	.90
Education (years)	3.06	1.00	3.14	1.01	0.33	.74
Gender (male/female)	26/7		26/9		0.19	.66
Diagnosis
With ASD^a	12		15		0.30	.58
With ADHD^a	21		20
Language
CVT total score	4.45	3.91	5.69	5.50	1.06	.29
RAN total correct	77.82	29.01	77.18^c	22.33	−0.10	.92
RAN total errors	0.85	1.56	0.85^c	1.16	0.01	.99
Working memory
NST-F score	10.03^b	2.65	10.63	3.18	0.83	.41
NST-B score	3.97^b	1.64	3.63	2.18	−0.72	.48
Sustained attention
DCT total time (s)	576.21	67.02	588.14	26.79	0.97	.33
DCT total score	138.33	36.04	140.31	43.21	0.21	.84
Learning and memory
HKLLT total learning	14.48	6.69	17.35^c	7.09	1.70	.09
HKLLT delayed recall	4.58	3.02	5.68^c	3.18	1.45	.15

Children diagnosed with ASD or ADHD include those with a single diagnosis and those with other comorbidities.

n = 32.

n = 34.

ASD: autism spectrum disorder; ADHD: attention-deficit/hyperactivity disorder; CLR: conventional language remediation; CVT: Chinese Vocabulary Test; DCT: Digit Cancellation Test; HKLLT: Hong Kong List Learning Test; NST-B: Number Span Test-Backward; NST-F: Number Span Test-Forward; RAN: Rapid Automatic Naming; SD: standard deviation; TPT: temporal processing tele-intervention.

Further analyses were performed to examine whether children with ASD and ADHD had different baseline characteristics and cognitive functions. Table 2 shows that two diagnostic groups had comparable age (t(66) = −0.85, p = .40) and gender (χ²(1) = 3.84, p = .05), but the ADHD group had a significantly higher level of education than the ASD group (t(66) = −2.73, p = .008). With education level as a covariate in ANCOVA, the two diagnostic groups showed similar performance levels on many cognitive measures (p > .05) as shown in Table 2, except that children with ASD performed better than those with ADHD in the CVT total score (p = .002), the RAN total correct score (p = .007) and the HKLLT total learning (p = .001).

Table 2.

Baseline characteristics and cognitive performance of the two diagnostic groups.

Baseline characteristics/cognitive performance	ASD (n = 27)		ADHD (n = 41)		t/χ²	p
Baseline characteristics/cognitive performance	M	SD	M	SD	t/χ²	p
Age (years)	8.32	0.88	8.54	1.17	−0.85	.40
Education (years)	2.70	0.78	3.37	1.09	−2.73	.008**
Gender (male/female)	24/3		28/13		3.84	.05
	M	SD	M	SD	F	p
Language
CVT total score	6.30	5.99	4.29	3.70	10.01	.002**
RAN total correct	83.52	25.21	73.43^a	25.43	7.67	.007**
RAN total errors	0.52	0.85	1.08^a	1.59	2.76	.10
Working memory
NST-F score	11.08^b	3.46	9.88	2.47	2.73	.10
NST-B score	4.08^b	2.26	3.61	1.70	3.99	.05
Sustained attention
DCT total time (s)	595.00	16.29	574.02	62.55	0.11	.74
DCT total score	130.70	35.36	145.05	41.61	0.00	.99
Learning and memory
HKLLT total learning	18.22	8.29	14.40^a	5.57	11.76	.001**
HKLLT delayed recall	5.89	3.71	4.63^a	2.59	2.70	.11

n = 40.

n = 26.

^**

p < .01.

ASD: autism spectrum disorder; ADHD: attention-deficit/hyperactivity disorder; CVT: Chinese Vocabulary Test; DCT: Digit Cancellation Test; HKLLT: Hong Kong List Learning Test; NST-B: Number Span Test-Backward; NST-F: Number Span Test-Forward; RAN: Rapid Automatic Naming; SD: standard deviation.

Intervention attendance and progress

The TPT group (M = 77.4%, SD = 20.1%) and the CLR group (M = 81.7%, SD = 21.2%) had a similar attendance rate (t(66) = −0.85, p = .40). No adverse side effects were reported in both groups. On average, children in the TPT group completed 85.6% (SD = 27.5) of the tone training and 33.7% (SD = 39.0) of the CV training. The overall completion mean of the two training modules was 59.6% (SD = 27.9). While 71% of children completed 100% of the tone training, only 17% completed 100% of the CV training. Relatively fewer children have completed the CV module because many demonstrated difficulty discriminating the CV stimuli with a shortened formant transition, especially for stimuli that sounded closer (e.g. /ka/ vs. /ta/, /da/ vs. /ga/). Given the three-down-one-up adaptive procedure, few children could achieve a 100% completion rate of the CV training.

Improvement in language after TPT

Children who received TPT demonstrated a trend of improvement in rapid naming accuracy, as measured by the RAN total errors (Table 3). Results of two-way time×group mixed ANOVA showed a marginally significant main effect of time (F(1,65) = 4.84, p = .03) with Bonferroni-adjusted alpha value, and the effect size was medium (η_p²= 0.07). Yet, there was a nonsignificant time by group interaction effect (F(1,65) = 0.04, p = .84). The marginally significant main effect of time suggests that children demonstrated a trend of fewer errors committed after both types of intervention. Though the multivariate results did not suggest a significant interaction effect, post hoc paired t-test showed that children who received TPT committed fewer rapid naming errors, which was marginally significant at Bonferroni-adjusted alpha value at .025 (t(33) = 2.04, p < .05). For the CLR group, the extent of error reduction was not statistically significant (t(32) = 1.37, p = .18). The present results suggest that children after receiving TPT seem to have a more apparent trend of committing fewer rapid naming errors than CLR.

Table 3.

Cognitive performance before and after CLR and TPT.

Cognitive measures	CLR (n = 33)							TPT (n = 35)
	Pre		Post		t	p	d	Pre		Post		t	p	d
	M	SD	M	SD	t	p	d	M	SD	M	SD	t	p	d
Language
CVT total score	4.45	3.91	7.18	5.20	−4.72	<.001***	0.82	5.69	5.50	8.49	6.41	−5.60	<.001***	0.95
RAN total correct	77.82	29.01	91.94	33.72	−3.93	<.001***	0.68	77.18^b	22.33	85.59^b	23.26	−4.38	<.001***	0.75
RAN total errors	0.85	1.56	0.42	0.75	1.37	.18	0.24	0.85^b	1.16	0.50^b	0.83	2.04	.05	0.35
Working memory
NST-F score	10.03^a	2.65	10.13^a	2.56	−0.39	.70	0.07	10.63	3.18	10.60	2.40	0.08	.94	0.01
NST-B score	3.97^a	1.64	4.53^a	2.11	−1.83	.08	0.32	3.63	2.18	5.17	2.48	−4.85	<.001***	0.82
Sustained attention
DCT total time (s)	576.21	67.02	566.36	61.02	1.27	.21	0.22	588.14	26.79	561.83	66.52	2.66	.012*	0.45
DCT total score	138.33	36.04	168.18	24.65	−6.09	<.001***	1.06	140.31	43.21	163.00	27.09	−4.00	<.001***	0.68
Learning and memory
HKLLT total learning	14.48	6.69	19.03	8.78	−4.47	<.001***	0.73	17.35^b	7.09	20.32^b	7.70	−3.13	.004**	0.52
HKLLT delayed recall	4.58	3.02	6.39	3.21	−4.63	<.001***	0.81	5.68^b	3.18	6.82^b	2.99	−3.11	.004**	0.53

n = 32.

n = 34.

p < .025 (Bonferroni-adjusted alpha level for measures from the same cognitive test); ^** p < .01; ^*** p < .001; d represents the effect size in terms of Cohen's d.

CLR: conventional language remediation; CVT: Chinese Vocabulary Test; DCT: Digit Cancellation Test; HKLLT: Hong Kong List Learning Test; NST-B: Number Span Test-Backward; NST-F: Number Span Test-Forward; RAN: Rapid Automatic Naming; SD: standard deviation; TPT: temporal processing tele-intervention.

Rapid naming speed, as measured by the RAN total correct score, also did not show a significant time×group interaction effect (F(1,65) = 2.00, p = .16), but demonstrated a significant main effect of time (F(1,65) = 31.07, Bonferroni-adjusted p < .001) with a large effect size (η_p²= 0.32). Both groups named significantly more letters within the time limits (i.e. at a faster naming speed; TPT: t(33) = −4.38, p < .001; CLR: t(32) = −3.93, p < .001; Table 3), and their extent of improvement was also comparable (t(65) = −1.41, p = .16).

Word knowledge, as measured by the CVT total score, did not show a significant time×group interaction effect (F(1,66) = 0.01, p = .92), but demonstrated a significant main effect of time (F(1,66) = 52.68, p < .001) with a large effect size (η_p²= 0.44). Both the TPT and CLR groups demonstrated significant improvement in word knowledge after intervention (TPT: t(34) = −5.60, p < .001; CLR: t(32) = −4.72, p < .001; Table 3). Compared to the CLR group, the TPT group showed a similar level of improvement in word knowledge (t(66) = 0.10, p = .92).

In summary, TPT and CLR significantly improve word knowledge and rapid naming speed of children with neurodevelopmental disorders. In addition, TPT has a more apparent positive effect on rapid naming accuracy.

Improvement in working memory after TPT

Two-way time×group mixed ANOVA showed a significant interaction effect on the NST-B score (F(1,65) = 4.87, Bonferroni-adjusted p = .03) with a medium effect size (η_p²= 0.07). The main effect of time (F(1,65) = 22.45, Bonferroni-adjusted p < .001) was also significant. The results suggest that after the intervention, the TPT group, but not the CLR group, demonstrated significantly higher NST-B scores (t(34) = −4.85, p < .001) with a large effect size (Table 3). The increase in the NST-B score in the TPT group (+ 1.54) was 2.8 times that in the CLR group (+ 0.56), and the between-group difference in the score increase was statistically significant (t(65) = 2.21, p = .03, d = 0.54; Figure 1).

Figure 1.

Performance changes on the Number Span Test-Backward score before and after the intervention. The pre–post difference score is computed as postmeasurement minus premeasurement, where a positive value indicates improved performance after the intervention The error bar represents the +/− 1 standard error of the mean. *** p < .001 indicates significant improvements in the NST-B score after TPT based on paired samples t-test with Bonferroni adjustment. * p < .05 indicates a significantly greater increase in the NST-B score after TPT compared to after CLR based on independent samples t-test.. CLR: conventional language remediation; NST-B: Number Span Test-Backward; TPT: temporal processing tele-intervention.

By contrast, there was no significant change in the NST-F score in any intervention group. The result of a mixed ANOVA did not show significant main and interaction effects for the NST-F score (F(1,65) = 0.02 and 0.08, p = .88 and .78, respectively). That means the positive effect of TPT is specific to the working memory that requires mental effort in manipulating auditory stimuli (i.e. NST-B) but not for simply holding these stimuli temporarily without mental manipulation (i.e. NST-F).

Improvement in sustained attention after TPT

The result of two-way time×group mixed ANOVA on the total completion time of the DCT did not show a significant time×group interaction effect (F(1,66) = 1.69, p = .20). Still, it demonstrated a significant main effect of time (F(1,66) = 8.14, Bonferroni-adjusted p = .006) with a medium effect size (η_p²= 0.11). The significant main effect of time suggests that children could complete the task in a shorter time after both types of intervention. Although the interaction effect was not significant, the results of paired t-tests showed that the TPT group showed significantly faster speeds in canceling the target digits after intervention (t(34) = 2.66, Bonferroni-adjusted p = .012) with an effect size approaching medium level (d = 0.45), whereas the CLR group did not (t(32) = 1.27, p = .21, d = .22; Table 3). After TPT, children became 26 seconds faster in completing the DCT, which was 2.7 times the reduction in completion time in children receiving CLR (10 seconds less). Therefore, the TPT group tends to have a more significant positive effect on processing speed than the CLR group (Figure 2).

Figure 2.

Performance changes in the measures of the Digit Cancellation Test before and after the intervention. The pre–post difference score is computed as postmeasurement minus premeasurement, where a negative value for the DCT total time and a positive value for the DCT total score indicates an improvement after intervention. The error bar represents the +/− 1 standard error of the mean. * p < .05 indicates a significant reduction in the DCT total time after TPT based on paired samples t-test with Bonferroni adjustment. *** p < .001 indicates a significant increase in the DCT total score after both types of interventions based on paired samples t-test. CLR: conventional language remediation; DCT: Digit Cancellation Test; TPT: temporal processing tele-intervention.

Moreover, children not only work faster but also better. The result of two-way time×group mixed ANOVA on the DCT total score showed a significant main effect of time (F(1,66) = 48.60, Bonferroni-adjusted p < .001) with a large effect size (η_p²= 0.42), though the time×group interaction effect was not significant (F(1,66) = 0.90, p = .35). The TPT group showed a significantly higher DCT total score after intervention (t(34) = −4.00, p < .001), and a similar improvement was also observed in the CLR group (t(32) = −6.09, p < .001; Table 3).

In summary, in a sustained attention task, the TPT and the CLR programs show similar significant improvement in accuracy. Although both groups of children complete the task in a shorter time, TPT, as compared to CLR, may have a higher impact on processing speed in children with neurodevelopmental disorders, given its significant pre–post difference after Bonferroni adjustment.

Improvement in learning and memory after TPT

Both the TPT and CLR groups showed improvements in verbal learning and memory after intervention (Table 3). For total learning and delayed recall measured with the HKLLT, the results of two-way time×group mixed ANOVA did not show a significant time×group interaction effect (total learning: F(1,65) = 1.23, p = .27; delayed recall: F(1,65) = 1.56, p = .22), but demonstrated a significant main effect of time (total learning: F(1,65) = 27.96, p < .001; delayed recall: F(1,65) = 30.37, p < .001) with Bonferroni adjustment. Children with neurodevelopmental disorders demonstrated significant improvement in verbal learning and memory after TPT (total learning: t(33) = −3.06, p = .004; delayed recall: t(33) = −3.11, p = .004) and CLR (total learning: t(32) = −4.37, p < .001; delayed recall: t(32) = −4.63, p < .001). The extents of their improvement in verbal learning (t(65) = −1.11, p = .27) and memory (t(65) = −1.25, p = .22) were comparable. Given that children were exposed to a learning environment on Chinese literacy in CLR, this may facilitate their ability to learn and memorize new Chinese words. However, children who received TPT were not provided with any direct intervention on learning and memory but still demonstrated significant enhancement in verbal learning and delayed recall to an extent similar to CLR.

Discussion

This study showed that children with ASD and ADHD demonstrated improvement in language, working memory, sustained attention, and verbal learning and memory through remote online training on their temporal processing skills. The improvement in working memory requiring mental manipulation of auditory stimuli was specific to TPT but not CLR. There was an overall improvement in sustained attention in terms of processing speed and rapid naming accuracy after the intervention; however, the statistically significant pre–post difference after receiving TPT but not CLR might suggest that TPT seemed to have a more apparent trend of improvement. Improvements due to TPT also extended to word knowledge, rapid naming speed, sustained attention in terms of accuracy, verbal learning, and memory to a degree comparable to those after CLR. These findings suggest that TPT has positive effects on various cognitive functions of children with ASD and ADHD and that the enhancement in working memory is specific to TPT. Therefore, remote online training on auditory temporal processing can be a potential intervention for enhancing the cognitive functions of children with ASD and ADHD.

While previous studies primarily provided in-person TPI,^24–29 the present study has added knowledge regarding the possibility and effectiveness of a web-based TPT that can be delivered remotely for children with neurodevelopmental disorders. Other studies have also reported the benefits of remote online intervention for children with neurodevelopmental disorders.^33–35 For instance, Aloizou et al.³³ found that 7 to 15-year-old children with autism who attended an online game-like educational program combined with a video conferencing platform for 1 to 2 months demonstrated more significant achievement in reaching their learning goals. Therefore, the web-based mode of intervention may provide a more flexible and extensive approach for reaching children with neurodevelopmental disorders, making intervention more accessible for children in lower-income classes or living in remote areas. The remote modality of TPT may be more cost-effective than in-person training, as the children can do the training independently or with limited assistance from adults at home. Therefore, it may reduce the intervention cost as the required staff will be reduced. Since TPT is less dependent on language and culture; therefore, it can be applied to children in different countries. In addition, the remote TPT allows children in less advantaged backgrounds (i.e. developing countries) to receive training that may benefit them.

The positive effects of auditory TPT on rapid naming and word knowledge found in the present study were in line with some previous studies.^24,27,28 For instance, Wang et al.²⁷ found that TPI exhibited significant correlations with rapid naming and corresponding reading-related abilities among Chinese children with dyslexia. A meta-analysis has reported a moderate-to-strong relationship between rapid automatized naming and reading performance.⁴⁷ Besides, the RAN test has been reported to be significantly loaded in both phonological processing and processing speed component in principal component analyses among Chinese dyslexic children.⁴⁸ Therefore, it is postulated that temporal processing training that targets improving perceptual processing in the time domain as short as in milliseconds may probably facilitate faster and more accurate letter naming in the RAN test. Tallal et al.²⁴ also reported improved speech discrimination and comprehension after TPI for children with a language-learning impairment, suggesting that children can process and understand others’ speech more accurately. It is postulated that more efficient tone and speech processing may facilitate children's better learning and understanding of new vocabulary. While the previous interventions were face-to-face, the present study further suggests that TPI can be conducted remotely and yields similar effects.

Apart from the language domain, the present findings are also consistent with previous studies that have reported enhancements in sustained attention,³⁰ working memory,^31,32,49 and verbal learning and memory.^31,32,49 However, the present study differs from those studies in terms of participants’ characteristics, intervention intensity, and the complexity of intervention modules. Many of these studies trained healthy older adults for 32 to 40 sessions,^30–32,49 4 to 5 days per week for 8 to 10 weeks, yet the present study offered less intensive intervention (20 weekly sessions for 8 months) to children with neurodevelopmental disorders. In addition, previous studies provided intervention programs comprising at least six to nine intervention modules,^30–32,49 spanning from simple acoustic tasks (e.g. TOJ of rapidly successive frequency-modulated sweeps) to complex manipulations of continuous speech (e.g. narrative memory). Hence, the positive outcomes are mixed results from different forms of TPI, and the unique contribution of TOJ-based intervention remains unknown.

The results of the positive effects of TPT on children's working memory and sustained attention are particularly encouraging. Children with ASD⁵⁰ and ADHD⁵¹ are commonly found to have impairment in working memory and sustained attention. In addition, the capacity of working memory has been reported to be a strong predictor of cognitive development in childhood,^52,53 learning outcomes in literacy and numeracy,^54–56 and overall academic performance.⁵⁷ Therefore, an appropriate and practical intervention that can enhance working memory may facilitate the learning of children with neurodevelopmental disorders and, in the long run, reduce the gaps in cognitive functions and academic performance between children with neurodevelopmental disorders and their peers.

Despite the encouraging findings, there are a few limitations to this study. First, this study only examined the immediate effects of TPT. Whether such positive outcomes can be maintained after intervention warrants future follow-up studies. Second, it is unknown whether cognitive enhancements after TPT and CLR are also related to natural neurodevelopment or learning through education for 8 months. Therefore, it is worthwhile to include a waitlist control group in future studies to explore this issue. Third, without applying neurophysiological measures in this study, the neural mechanisms underlying the positive effects of TPT remain unknown. Thus, further studies applying technologies such as electroencephalogram or functional near-infrared spectroscopy can be conducted to shed some light on intervention-related neural responses. Last but not least, the present study did not include an additional outcome measure of temporal processing ability per se; thus, how the improvement in temporal processing might contribute to the enhancement of cognitive functions remains inconclusive. Therefore, future studies may include different temporal processing measures and examine their association with training-related cognitive performance changes.

Conclusions

Temporal processing deficits found in ASD and ADHD have been proposed to be a possible factor mediating their cognitive impairment in attention, memory, and language. While in-person TPI was reported to improve various cognitive functions in children, the present study has found similar positive effects of remotely delivered web-based TPT on children with ASD and ADHD. Children after receiving TPT, but not CLR, demonstrated a specific enhancement in working memory. The TPT group shows a greater impact on rapid naming accuracy and sustained attention in processing speed than the CLR group. In addition, the TPT group showed significantly improved word knowledge, rapid naming speed, sustained attention in terms of accuracy, and verbal learning and memory, to an extent similar to that of the CLR group. The present findings suggest that web-based TPT may be a potential intervention for improving cognitive functions in children with neurodevelopmental disorders.

Footnotes

Acknowledgements

The authors would like to thank Quin Chan, Christy Cheung, Kenneth Lai, Katie Leung, Evan Lau, Tsz-Lok Lee, Camelia Ding, and Natalie Auyeung for collecting data and/or preparing and assisting with the group intervention for this study. Further appreciation is extended to the school principals and teachers for their recruitment of participants and all forms of support throughout the experiment, as well as to all children and parents for their participation.

Contributorship

ASC contributed to the conception and design of the study. SLS contributed to data acquisition and management. ASC and SLS contributed to the formal analysis of data and drafting of the manuscript. All authors contributed to the interpretation of the data and the review and editing of the manuscript. All authors approved the final version of the manuscript.

ORCID iDs

Agnes S. Chan

Sophia L. Sze

Mei-Chun Cheung

Data availability

The dataset generated and analyzed during the current study is available from the corresponding author upon reasonable request.

Declaration of conflicting interests

The author(s) declared the following potential conflicts of interest with respect to the research, authorship, and/or publication of this article: The first author (ASC) is the founder of Pro-Talent Association Ltd. The other authors have no conflicts of interest to disclose.

Ethical approval

The Joint Chinese University of Hong Kong—New Territories East Cluster Clinical Research Ethics Committee approved this study (CREC reference number: 2020.501-T).

Funding

The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work was supported by the Lee Hysan Foundation Ltd.

Guarantor

ASC

References

Foster

Kisley

Davis

, et al. Cognitive function predicts neural activity associated with pre-attentive temporal processing. Neuropsychologia 2013; 51: 211–219.

Nobre

Van Ede

. Anticipated moments: temporal structure in attention. Nat Rev Neurosci 2018; 19: 34–48.

Paton

Buonomano

. The neural basis of timing: distributed mechanisms for diverse functions. Neuron 2018; 98: 687–705.

Pöppel

. A hierarchical model of temporal perception. Trends Cogn Sci 1997; 1: 56–61.

Tallal

Piercy

. Defects of non-verbal auditory perception in children with developmental aphasia. Nature 1973; 241: 468–469.

Chan

Ding

Lee

, et al. Temporal processing deficit in children and adolescents with autism spectrum disorder: an online assessment. Digital Health 2023; 9: 20552076231171500.

De Boer-Schellekens

Eussen

Vroomen

. Diminished sensitivity of audiovisual temporal order in autism spectrum disorder. Front Integr Neurosci 2013; 7: 8.

Kwakye

Foss-Feig

Cascio

, et al. Altered auditory and multisensory temporal processing in autism spectrum disorders. Front Integr Neurosci 2011; 4: 129.

Meilleur

Foster

Coll

, et al. Unisensory and multisensory temporal processing in autism and dyslexia: a systematic review and meta-analysis. Neurosci Biobehav Rev 2020; 116: 44–63.

10.

Breier

Gray

Fletcher

, et al. Perception of speech and nonspeech stimuli by children with and without reading disability and attention deficit hyperactivity disorder. J Exp Child Psychol 2002; 82: 226–250.

11.

Cardy

Tannock

Johnson

, et al. The contribution of processing impairments to SLI: insights from attention-deficit/hyperactivity disorder. J Commun Disord 2010; 43: 77–91.

12.

Chan

Ding

Lee

, et al. Temporal processing deficit in children with attention-deficit/hyperactivity disorder: an online assessment. Digital Health 2022; 8: Article 20552076221120325.

13.

Fostick

. The effect of attention-deficit/hyperactivity disorder and methylphenidate treatment on the adult auditory temporal order judgment threshold. J Speech Lang Hear Res 2017; 60: 2124–2128.

14.

Poole

Gowen

Warren

, et al. Brief report: which came first? Exploring crossmodal temporal order judgements and their relationship with sensory reactivity in autism and neurotypicals. J Autism Dev Disord 2017; 47: 215–223.

15.

Turi

Karaminis

Pellicano

, et al. No rapid audiovisual recalibration in adults on the autism spectrum. Sci Rep 2016; 6: 21756.

16.

West

Douglas

Houghton

, et al. Time perception in boys with attention-deficit/hyperactivity disorder according to time duration, distraction and mode of presentation. Child Neuropsychol 2000; 6: 241–250.

17.

Radonovich

Mostofsky

. Duration judgments in children with ADHD suggest deficient utilization of temporal information rather than general impairment in timing. Child Neuropsychol 2004; 10: 162–172.

18.

Allman

Pelphrey

Meck

. Developmental neuroscience of time and number: implications for autism and other neurodevelopmental disabilities. Front Integr Neurosci 2012; 6: Article 7.

19.

Maister

Plaisted-Grant

. Time perception and its relationship to memory in autism spectrum conditions. Dev Sci 2011; 14: 1311–1322.

20.

Foss-Feig

Schauder

Key

, et al. Audition-specific temporal processing deficits associated with language function in children with autism spectrum disorder. Autism Res 2017; 10: 1845–1856.

21.

Hart

Marquand

Smith

, et al. Predictive neurofunctional markers of attention-deficit/hyperactivity disorder based on pattern classification of temporal processing. J Am Acad Child Adolesc Psychiatry 2014; 53: 569–578.e1.

22.

Noreika

Falter

Rubia

. Timing deficits in attention-deficit/hyperactivity disorder (ADHD): evidence from neurocognitive and neuroimaging studies. Neuropsychologia 2013; 51: 235–266.

23.

Hendrich

Strobach

Buss

, et al. Temporal-order judgment of visual and auditory stimuli: modulations in situations with and without stimulus discrimination. Front Integr Neurosci 2012; 6: Article 63.

24.

Tallal

Miller

Bedi

, et al. Language comprehension in language-learning impaired children improved with acoustically modified speech. Science 1996; 271: 81–84.

25.

Rey

De Martino

Espesser

, et al. Temporal processing and phonological impairment in dyslexia: effect of phoneme lengthening on order judgment of two consonants. Brain Lang 2002; 80: 576–591.

26.

Habib

Rey

Daffaure

, et al. Phonological training in children with dyslexia using temporally modified speech: a three-step pilot investigation. Int J Lang Commun Disord 2002; 37: 289–308.

27.

Wang

Liu

. Distinct effects of visual and auditory temporal processing training on reading and reading-related abilities in Chinese children with dyslexia. Ann Dyslexia 2019; 69: 166–185.

28.

Strehlow

Haffner

Bischof

, et al.

Does successful training of temporal processing of sound and phoneme stimuli improve reading and spelling?

Eur Child Adolesc Psychiatry 2006; 15: 19–29.

29.

Stevens

Fanning

Coch

, et al. Neural mechanisms of selective auditory attention are enhanced by computerized training: electrophysiological evidence from language-impaired and typically developing children. Brain Res 2008; 1205: 55–69.

30.

Szelag

Skolimowska

. Cognitive function in elderly can be ameliorated by training in temporal information processing. Restor Neurol Neurosci 2012; 30: 419–434.

31.

Mahncke

Bronstone

Merzenich

. Brain plasticity and functional losses in the aged: scientific bases for a novel intervention. Prog Brain Res 2006; 157: 81–109.

32.

Mahncke

Connor

Appelman

, et al. Memory enhancement in healthy older adults using a brain plasticity-based training program: a randomized, controlled study. Proc Natl Acad Sci 2006; 103: 12523–12528.

33.

Aloizou

Chasiotou

Retalis

, et al. Remote learning for children with special education needs in the era of COVID-19: beyond tele-conferencing sessions. Educ Media Int 2021; 58: 181–201. doi:10.1080/09523987.2021.1930477

34.

Kim

Fienup

. Increasing access to online learning for students with disabilities during the COVID-19 pandemic. J Spec Educ 2022; 55: 213–221.

35.

Panagouli

Stavridou

Savvidi

, et al. School performance among children and adolescents during COVID-19 pandemic: a systematic review. Children 2021; 8: Article 1134.

36.

Tallal

. Fast ForWord®: the birth of the neurocognitive training revolution. Prog Brain Res 2013; 207: 175–207.

37.

Ebaid

Crewther

. The contribution of oculomotor functions to rates of visual information processing in younger and older adults. Sci Rep 2020; 10: 10129.

38.

Chan

Cheung

Sze

, et al. Measuring vocabulary by free expression and recognition tasks: implications for assessing children, adolescents, and young adults. J Clin Exp Neuropsychol 2008; 30: 892–902.

39.

Working Group on Child Neuropsychology, Hong Kong Neuropsychological Association. Neuropsychological test battery for primary school Cantonese-speaking children in Hong Kong. Hong Kong: Hong Kong Neuropsychological Association, 2014.

40.

Chan

. Hong Kong list learning test. 2nd ed. Hong Kong: The Chinese University of Hong Kong, 2006.

41.

Hugdahl

. Dichotic listening in the study of auditory laterality. In: Hugdahl

Davidson

(eds) The asymmetrical brain. Cambridge, MA: The MIT Press, 2002, pp.441–475.

42.

Audacity Team. Audacity, https://www.audacityteam.org/ (2023, accessed 1 September 2019).

43.

Boersma

Weenink

. Praat: doing phonetics by computer, http://www.praat.org/ (accessed 1 September 2019).

44.

Hong Kong Specific Learning Difficulties Research Team. Intervention resources, https://hksld.eduhk.hk/Ourresources/Intervention%20Resources (2021, accessed 1 September 2020) (in Chinese).

45.

Hong Kong Institute of Education. Thematic network on support for diverse learning needs (reading and writing). Hong Kong: Department of Special Education and Counselling, The Hong Kong Institute of Education, 2012 (in Chinese).

46.

Education Bureau. Resource package on narrative skills and utterance expansion, https://sense.edb.gov.hk/en/types-of-special-educational-needs/speech-and-language-impairment/resources/teaching-resources/32.html (2021, accessed 1 September 2020) (in Chinese).

47.

Araújo

Reis

Petersson

, et al. Rapid automatized naming and reading performance: a meta-analysis. J Educ Psychol 2015; 107: 868–883.

48.

Pan

Shu

. Rapid automatized naming and its unique contribution to reading: evidence from Chinese dyslexia. In: Chen

Wang

Luo

(eds) Reading development and difficulties in monolingual and bilingual Chinese children. Literacy studies. Vol.8. Dordrecht: Springer, 2014, pp.125–138.

49.

Smith

Housen

Yaffe

, et al. A cognitive training program based on principles of brain plasticity: results from the Improvement in Memory with Plasticity-based Adaptive Cognitive Training (IMPACT) study. J Am Geriatr Soc 2009; 57: 594–603.

50.

Kercood

Grskovic

Banda

, et al. Working memory and autism: a review of literature. Res Autism Spectr Disord 2014; 8: 1316–1332.

51.

Ramos

Hamdan

Machado

. A meta-analysis on verbal working memory in children and adolescents with ADHD. Clin Neuropsychol 2020; 34: 873–898.

52.

Case

. Intellectual development from birth to adulthood: a neo-Piagetian interpretation. In: Siegler

(ed.) Children's thinking: what develops? 1st ed. Mahwah, NJ: Lawrence Erlbaum Associates, 1978, pp.37–71.

53.

Pascual-Leone

. A mathematical model for the transition rule in Piaget's developmental stages. Acta Psychol 1970; 32: 301–345.

54.

Alloway

Gathercole

Kirkwood

, et al. The cognitive and behavioral characteristics of children with low working memory. Child Dev 2009; 80: 606–621.

55.

Daneman

Carpenter

. Individual differences in working memory and reading. J Verbal Learn Verbal Behav 1980; 19: 450–466.

56.

Jarrold

Bayliss

. Variation in working memory due to typical and atypical development. In: Conway

ARA

Jarrold

Kane

Miyake

Towse

(eds) Variation in working memory. New York: Oxford University Press, 2007, pp.134–161.

57.

Alloway

. Investigating the predictive roles of working memory and IQ in academic attainment. J Exp Child Psychol 2010; 106: 20–29.