Table 1 Study characteristics of the selected articles

Stiles et al. [16]

Comparative study

To study the disturbances

in WM, and its relation to

the receptive vocabularies in mild to moderately

severe SNHL children.

Experimental group:

16 children (mild to severe SNHL) fitted with bilateral hearing aids.

Mean age: 3;92 [years; months])

Control group: 24 normal-hearing children

(13 boys, 11 girls)

Mean age: 9;94 [years;months)

FDS

BDS

FDS task:

Children had to repeat the digits in the same order as presented. Numbers were presented in every 1-s interval.

The digits were not repeated within the series. Additionally, “7” was omitted from the presentation to make all of them to be monosyllabic in nature. Children repeated two strings at each length. The three-digit sequence was tested first. The number of digits per sequence increased by one digit if any of the 3-digit strings could be correctly repeated. This gradual lengthening of the sequence persisted until either the children reached the string limit of 8 digits, the examiner stopped the process, or both strings of a single length were wrongly repeated back.

BDS task:

Children had to repeat the digits in reverse order of presentation. Except for starting with strings of two digits rather than three, test administration was the same as for FDS. The BDS task was administered similarly to the FDS task.

Four different conditions were used to administer the FDS and BDS: auditory quiet, auditory noise, visual quiet, and visual noise

HI children and CNH showed

an auditory advantage in forward span. HI children showed a similar memory span as that of CNH.

The presence of background noise did not affect performance in either group.

Children with moderate to moderately severe HL displayed a resilient WM system. Relationships between WM and vocabulary were observed for all children; those with poor WM demonstrated a smaller vocabulary size.

The presence of noise did not affect the performance of the digit task. This could be explained by the fact that either the background noise wasn't distracting enough to need a change in the executive resource allocation for the WM activities, or the resources required to decode the signal in noise are unrelated to those used for WM tasks.

Harris et al [10]

Longitudinal study

To investigate how verbal STM and WM function as a developmentally limiting source of variability in children’s speech and language outcomes longitudinally following CI.

Experimental group:

66 children with CI

Age range: 7–15 years

(Mean/SD not specified)

WISC-III Edition

The test involves the child repeating progressively larger lists of digits presented by the experimenter using live voice. Roughly one digit per second was presented.

There are two recall conditions in the task:

• DSF

• DSB

In order to complete the DSF task, participants must repeat a series of random digits, starting with a two-digit sequence and going in order from 1 to 9 (inclusive). For each sequence length, two things are presented, and if the subject reproduces at least one of them correctly, the sequence length is increased by one. This process continues until the participant repeats both of the wrong items at the same sequence length. The only difference between the DSB and DSF tasks is that subjects must reproduce the sequences in reverse order.

Compared with the normative mean scores (Population means: 10; SD: 3), the CI children’s scores were 1 SD below the normative in 50.5% in DSF and 44.0% in DSB across all ages.

However, the DSF and DSB performance slopes in the CI group that represented the development of verbal STM/WM capacity were comparable in magnitude to values found in a sample of people with normal hearing using WISC-III cross-sectional norms.

Variation in STM/WM is one of the fundamental neurocognitive-related factors that underlie all behavioral measures of S/L performance.

Soleymani et al. [15]

Cross-sectional study

To investigate WM as a cognitive ability in children with ND and CI.

Experimental group:

50 children with CI.

Age range: 5–7 years

Mean: 6.16

SD: 0.79

Control group:

50 children with ND.

Age range: 57 years

Mean: 6.16

SD: 0.79

NWR

FDS

BDS

Non-word repetition task

Two practice non-words were given to the children before moving on to the actual testing. Children were presented with the original target non-words, which they had to repeat. They received a score of 1 if they repeated the target non-word perfectly; otherwise, they received a score of 0. The total score for this task was 25.

FDS and BDS tasks

The child was told to repeat seven sets of numbers, 3–9 for FDS and 2–8 for BDS. In FDS, children were tasked with repeating the numbers in the exact same sequence as before. Children should just tell numerals in reverse order in BDS, though. Each number in the series was stated with a one-second interval. Every group of numbers is repeated twice. Each response was worth one point when it was correct. The task was concluded after two attempts at each sequence failed. The maximum score for each task was 14. The FDS was finished before the BDS.

Mean and SD

• FDS

NH: 5.42 (1.63)

CI: 2.30(1.43)

• BDS

NH: 3(1.95)

CI: 0.84(0.84)

• NWR

NH: 22.78(1.59)

CI: 12.28(1.82)

The NWR scores of the CI children showed a strong relationship with the FDS and BDS. Also, FDS and BDS are strongly associated. Although FDS and BDS were substantially correlated, the NWR scores of the ND children were shown to be only modestly related to them.

The working memory of children with CI has been compromised. The existence of differences between ND and CI children suggested that early exposure to sound had a significant impact on the part of the brain that stores and retains phonological information in STM. Children with CI were found to have similar developmental patterns. In the BDS, there was no discernible difference between preschool and first grade. Children with CIs may not perform as well on BDS in the early stages of development as with ND since it is a test designed to evaluate complex memory spans.

Tao et al. [18]

Correlational study

To study the relationship between AWM and speech perception performance in Mandarin-speaking CI children.

Experimental group:

32 CI users (21–pre-lingual HI

11–post-lingual HI)

Age range: 6.0–26.0 years

Mean: 13.0

SD: 4.0

Control group:

21 normal-hearing children

Age range: 8–14 years

Mean: 11.0

SD: 1.6

Auditory digit span test

An adaptive (1-up/1-down) approach was used to test auditory digit span recall in both the forward and backward directions. The stimuli featured a single man talker uttering the numbers 0 through 9. Numbers were chosen and delivered in random order (no visual cues). Children replied by clicking on the response boxes displayed on a computer screen in the order of the sequence of digits they heard. Three digits made up the first series. The number of digits provided was either increased or decreased depending on how accurately the response was given.

The mean score

CI group:

BDS: 4.72 (SE = 0.33)

FDS: 6.10 (SE = 0.35) for.

NH group:

BDS: 5.96 (SE = 0.30)

FDS: 7.39 (SE = 0.21)

Performance ranged from 1.8 to 11 for FDS and from 2.1 to 9.7 for backward digit span, indicating a significant inter-subject variability.

Only sentence recognition in quiet environments showed a significant correlation with AWM efficiency. The relationship between WM and lexical tone recognition was not observed.

Compared to NH participants, CI individuals' scores on the forward and backward digit span were significantly lower. Despite some similarity in the distributions of digit span scores, Mandarin-speaking CI individuals may be less capable and efficient at processing phonological information than NH participants, which would show up in their digit span. The connection between WM and speech performance was unaffected by pitch cues.

Torppa et al. [19]

Longitudinal study

To investigate the effects of musical experience, auditory working memory, and F0 auditory discrimination in the CI group.

Experimental group:

21 unilaterally implanted children (CIm and CIn group)

Control group:

21 normal-hearing children.

All the children were aged 4–13 years

FDS

ITPA FDS task was employed. In the analysis, raw scores were employed; these don't necessarily indicate how many digits are repeated.

Mean scores and SD in FDS task:

NH-T1 = 22.43 (7.10)

NH–T2 = 25.43 (7.72)

CIm-T1 = 20.38 (7.61)

CIm-T2 = 24.38 (9.52)

CIn-T1 = 15.69 (5.30)

CIn-T2 = 16.77

(5.83)

FDS in Cim and NH group were similar, but CIn group performed more poorly.

Similar findings were seen with respect to F0 discrimination and prosodic perception.

Children with CIs who have musical practice do better on the FDS than children without musical experience. Digit span and intensity perception were correlated to prosodic perception.

Willis et al. [20]

Longitudinal study

To compare verbal and visual WM of six children with congenital HI.

Experimental group:

6 children/adolescents with a unilateral cochlear implant

Age range: 8 and 15 years

Subtests of WMTB-C

Children are asked to recall words in the same order as presented in the word recall subtest. The NWR followed the same procedure, and the items followed a similar CVC structure as the “real” word. Four targets must be correctly recalled by the children. The difficulty of the tasks increases for each subgroup starting at two, three, or four. The subtest is terminated when a child is unable to correctly reproduce four targets from a subset.

Mean standard scores:

first-year of study

Word recall: 81.67 (SD 12.48)

Non-word recall: 110 (SD 10.66)

Second-year of study

Word recall: 80 (SD 13.19)

Non-word recall: 108 (SD 11.28)

Performance in non-word recall was better than word recall at both points of measurement.

For 2 years, all the children displayed the same pattern of verbal working memory. It was found that the children had more trouble with non-word memory than they did with word recall. Children with HI may not have appropriate phonological representations in their STM, which would make storing and retrieving information more challenging.

Visual WM was comparable to that of peers with average hearing.

AuBuchon et al. [1]

Longitudinal study

To study whether early-implanted, long-term CI users exhibit delays in verbal STM and WM capacity when audibility and speech production processes are excluded.

Experimental group: 23 CI users

The age range at initial testing: 7.8–15.3 years (Mean 11.8; SD 1.9)

Age at follow-up: 10.1–17.1 years (Mean 14.0; SD 2.4)

Control group: 23 NH controls

Age range at initial testing 8.2–15.3 years (Mean 12.5; SD 2.2)

Age range at follow-up: 10.1–16.6 years (Mean 14.0; SD 2.1)

WISC-III subsets

Lists of the set of digits were presented using live voice as per the instructions for the forward and backward span subtests of the WISC-III. The lists were to be repeated aloud by the participants in either forward or reverse order.

Forward span measures:

A significant effect of hearing status was seen across forward-span measures.

CI group mean 6.47 (SD 2.72)

NH group mean 8.3 (SD 3.25)

Significant effects on tasks were also seen.

ADS-F mean 8.25, SD 2.76

No interaction between hearing status and task.

NH group had a superior performance than CI users on all forward tasks.

Backward span measures:

No effect on hearing status was seen

Effect of task was seen

ADS-B mean 5.18, SD 2.14

No interaction between hearing status and tasks.

The performance on ADS-F had a strong positive correlation with a visual and computerized version of the digit span-forward task (no auditory stimuli were given) in long-term CI users.

While ADS-B did not correlate with the other 2 tasks. Also, the performance in ADS-B did not vary between the NH and CI groups as there was an increase in demand for processing the instructions.

Bharadwaj et al. [3]

Cross-correlational study

To study the WM and STM skills in the auditory and visual modalities in school-going CI children.

Study the relationship between verbal and visual WM/STM measures versus reading.

Experimental group:

10 children

Age 7 to 11 years (Mean/SD not specified)

CI users

WJ III COG NU

KABC-II

In the Numbers Reversed task, the test subject responds to a series of numbers before repeating them in a reverse manner. AWM is a task that assesses a person's capacity to hold a list of words and numbers in immediate awareness and then reorganize the information so that the words are remembered first and subsequently the numbers.

The Number Recall task tests a subject’s ability to retain auditory short-term information by having them listen to a set of numbers before having them repeated in the very same order. The Word order task requires the participant to touch the object’s silhouettes after hearing their names in the same order.

The mean standard scores on WJ III COG NU (numbers reversed and AWM) were less than the average SD (scores were < 85 in both numbers reversed and AWM).

Reading measures and the number recall test showed a strong positive association (except for passage comprehension and oral reading).

All CI children exhibited less than average performance in tasks related to verbal knowledge (number recall, word order) in KABC II. The outcomes are consistent with the idea that early-onset hearing loss impacts capacities like memory and creating sequential information due to sensory deprivation.

The abilities were related to the following:

Word reading abilities–Auditory STM

Passage comprehension abilities–Visual and Auditory WM.

Talebi, S., and Arjmandnia, A. A. [17]

Non-experimental, correlational, and causal-comparative study.

To study the interaction between WM and STM performance and their impact on CI outcomes

Experimental group:

31 CI children

Mean age: 121.52 months; SD: 19.946

Control group:

31 NH children

Mean age = 120.68 months;

SD = 18.137

WMTB-C

Not explained in detail.

Working memory scores (Mean (SD)):

NH: 68.9 (10.67)

HI 57.13 (8.64)

Working memory span:

NH 4.09 (0.65)

HI 3.61 (0.48)

Short-term memory score

NH 56.13(6.96)

HI 47.10 (6.57)

Short-term memory span:

NH 4.63 (0.62)

HI 3.87 (0.55)

WM efficiency was significantly and positively correlated with CI results. In addition, the children with cochlear implants performed worse than their counterparts with normal hearing.

Children with and without cochlear implants were found to have WM and STM that interacted well with each other. Working and short-term memories are enhanced in people implanted at younger ages.

There is a substantial correlation between AWM and STM, as well as between AWM, STM, and speech intelligibility.

Javanbakht et al. [12]

Cross-sectional study

To compare the memory abilities of two groups of children who used hearing aids in both ears and only differed in their capacity to understand speech in noisy environments.

Experimental group:

31 hearing aid user students

Participants were split into 2 groups:

• HP-BKB SIN score Less than or equal to 7

• LP- BKBSIN score greater than 7

Age range: 8–12 years were selected

Mean and SD 9.13 ± 0.17

FDS

BDS

NWR (Persian version)

FDS: The number series started out with two digits and eventually progressed to seven digits. The number of correctly memorized series is how the performance is measured. The exam was terminated after the child stated two wrong series.

BDS: Similar to that of FDS, except the subject must repeat the numbers backward.

Non-word repetition test:

40 non-sense syllable words. Each non-word contained 2, 3, or 4 syllables, and the interval between each presentation was 10 s or shorter, depending on how quickly the participant repeated each item. The test was carried out with a human voice and was based on the phonetically transcribed non-words that were accessible. The child hears the non-sense words read aloud, and he or she must repeat them. The number of correctly repeated non-words is used to grade performance.

Mean WM scores:

FDS:

HP-2.00 (± 0.50)

LP-1.52 (± 0.65)

BDS:

HP-2.00 (± 0.50)

LP-1.52 (± 0.68)

NWR:

HP-28.77 (± 5.04)

LP-22.33 (± 4.21)

The LP group had poor scores on FDS, BDS as well as NWR tests.

Speech in noise tests was significantly correlated with all the WM measures. This correlation was higher for NWR than for FDS and BDS.