A comment on the order of report in two-pair dichotic digit testing in children assessed for auditory processing

Abstract

Objective

This clinical note aimed to 1) describe the order-of-report (OoR) strategies used by a large sample of children who completed 2pDD testing in the free-recall condition as part of an audiological assessment of auditory processing (AP), and 2) determine if use of OoR correlated with 2pDD test performance.

Design

A retrospective (case-review), single observation design.

Study sample

A convenient sample of 77 children (50 males and 27 females aged 7.0 to 17.4 years [median = 8.8 years]) referred to audiology for an AP assessment.

Results

The participating children used temporal more than ear or sequential OoR and showed decreased test scores with increased use of ear OoR.

Conclusions

These findings suggest children referred for an AP assessment could favour a temporal OoR when completing 2pDD testing for its opportunity to rehearse the first digit pair (possibly strengthening its memory trace) prior to processing the second digit pair, while their use of an ear OoR could come at the cost detrimental ear dominance effects that reduce their test performance. Future research could consider if similar results are observed in other child populations and if preferred dichotic processing strategy could inform auditory training for children with dichotic listening deficits.

Keywords:

• Two-pair dichotic digits (2pDD)
• free-recall
• order-of-report (OoR)
• auditory processing disorder
• children

The two-pair dichotic digits test in the free-recall condition remains widely used in the audiological assessment of auditory processing (AP) in children around the world (AAA 2010; ASHA 2005; BSA 2011; Emanuel, Ficca, and Korczak 2011; Ismen and Emanuel 2023; Keith et al. 2019; Musiek and Weihing 2011; Weihing et al. 2015). In the free recall condition, the 2pDD is thought to assess the child’s ability to binaurally integrate disparate auditory stimuli (in this case, different digits) presented to each ear simultaneously (AAA 2010; ASHA 2005). While 2pDD testing is commonly used by audiologists to assess AP, the influence of cognitive processes are noted with the child needing to sufficiently attend to the digits and remember all four test items long enough to provide a complete response to the tester (Cameron and Dillon 2020; Cameron, Glyde, Dillon, and Whitfield 2016; Cameron, Glyde, Dillon, Whitfield, and Seymour 2016; Westerhausen and Hugdahl 2008).

Current recordings of 2pDD tests in the English-language typically consist of a pre-recorded list of 20 to 25 stimulus presentations (Cameron, Glyde, Dillon, Whitfield, and Seymour 2016; Musiek 1983). Each presentation contains four different numbers from 1 to 10 (excluding 7). These four digits are presented in two sequential pairs approximately one second apart, with each pair presenting a different digit to each ear simultaneously. In the free-recall condition, the listener is instructed to wait until both digit pairs are presented before reporting all four digits in any order. Responses are tallied to generate right ear, left ear, and ear advantage scores for comparison against age-appropriate normative data (AAA 2010; ASHA 2005; BSA 2011; Emanuel, Ficca, and Korczak 2011; Ismen and Emanuel 2023; Keith et al. 2019; Moncrieff 2011; Moncrieff and Musiek 2002; Musiek and Weihing 2011; Weihing et al. 2015).

Missing from current analyses of 2pDD testing in children is a consideration of order of report (OoR): the order in which the participant reports the 2pDD stimuli. Early research into DD testing in adults reported three broad OoR strategies (Broadbent 1954; Bryden 1962, 1964; Carr 1969; Parkinson et al. 1974; Satz et al. 1965; Rollins, Schurman, and Everson 1972; Savin 1967). The first was ear OoR where digits were reported by ear of presentation. The second was temporal OoR where digits were reported by temporal order of presentation. The third was sequential OoR where digits were reported in sequence as one digit each from the first and subsequent digit pairs followed by the other digit from the first and subsequent digit pairs.

Only three studies have considered OoR in DD testing of children (Neufeldt 1966; Surwillo et al. 1972; Witelson and Rabinovitch 1971). These studies assessed children on DD test stimuli ranging from 1- to 5-pairs presented at 1 pair/0.25 s to 1 pair/2 s. Whilst only considering ear and temporal OoR (these studies did not explicitly consider sequential OoR), general findings in typically developing children included a preference for ear OoR as presentation rate increased, a preference for temporal OoR as number of digit pairs increased, and an ability to change OoR strategy even within the same test session. These results were thought to reflect differences in how children prefer to encode or store dichotically presented information and raised the possibility that some OoR strategies could be more effective than others for processing dichotic stimuli.

To the authors’ knowledge, OoR has not been investigated in 2pDD testing in children as currently used in the audiological assessment of AP. This clinical note therefore aimed to 1) describe the OoR used by a large sample of children who completed 2pDD testing in the free-recall condition as part of an audiological AP assessment, and 2) determine if use of OoR correlated with 2pDD test performance.

Method

Research design

A retrospective (case-review), single observation design was used for this clinical note.

Participants

The participants were 77 school-aged children (50 males and 27 females; median age 8.8 years with minimum = 7.0, 1^st quartile = 7.8, 3^rd quartile = 11.2, and maximum = 17.4 years), conveniently sampled from children attending the Audiology Clinic of a large university in Queensland, Australia, between May 2020 and December 2021. These participants: a) had been referred for an auditory processing (AP) assessment by a wide range of persons including parents, teachers and health care professionals; b) were first language speakers of Australian English with no diagnoses of intellectual, cognitive, attentional, emotional or articulation impairments (as reported by their parent/guardian: not formally assessed); c) were attending a regular public or private school in Australia; f) showed no evidence of peripheral hearing loss (i.e. pure tone thresholds ≤15 dB HL at octave frequencies from 250 Hz to 8000 Hz, speech performance-intensity functions within normal limits as per Travers [1990]) and tympanograms with a static compliance ≥0.2 ml and tympanic peak pressure between −100 and +50 daPa as per Jerger [1970]); g) had completed at least four tests from the behavioural AP test battery used at the audiology clinic where this clinical note was conducted with one of those tests being a 2pDD test in the free recall condition, and h) had their parents/guardians consent for their 2pDD test results to be used for research. No restrictions were placed on the participants with regards to a diagnosis of auditory processing disorder (APD).

Two-pair dichotic digits testing

The 2pDD testing was conducted in sound-treated booths using commercially available, calibrated audiometric equipment and the Wilson and Strouse (1998) 2pDD test recording. All stimuli were presented under headphones at 50 dB HL (dial) with an approximately 1.1 s interval between the onset of the first and the second digit pairs (resulting in a presentation rate of approximately 1 pair/1.1 s). All responses were in the free-recall condition. Correct responses were scored as one point (2.5%) per digit correctly repeated per ear for 20 test stimuli.

Data collection and analysis

Participant responses to the 2pDD stimuli were scored for: 1) total digits correctly reported, 2) total items where all four digits were correctly reported, 3) total items where both digits presented to the right ear were correctly reported, 4) total items where both digits presented to the left ear were correctly reported, 5) the ear advantage score calculated as score 3 minus score 4 above, 6) total items where the first digit presented to the right ear was correctly reported, 7) total items where the second digit presented to the right ear was correctly reported, 8) total items where the first digit presented to the left ear was correctly reported, and 9) total items where the second digit presented to the left ear was correctly reported.

Each participant’s OoR for each 2pDD stimulus was categorised into one of seven types: 1) temporal forward – digits from the first pair were reported before digits from the second pair, 2) temporal backward – digits from the second pair were reported before digits from the first pair, 3) right ear – digits from the right ear were reported before digits from the left ear, 4) left ear – digits from the left ear were reported before digits from the right ear, 5) sequential – a digit from the first pair and a digit from the second pair were reported before the other digit from the first pair and the other digit from the second pair, 6) ear or sequential – the OoR given fits both ear and sequential orders of report in a manner that prevents exclusive allocation to either, or 7) no identifiable pattern – the OoR did not match any of the above. Each participant’s use of each OoR strategy was tallied and converted to proportions of use for that participant’s 2pDD test assessment.

Two sets of inferential analyses were conducted on the 2pDD test scores and OoR use by the participant group. The first set of analyses used one-sample Wilcoxon signed rank tests at the 5% level to compare within participant use of temporal OoR strategies (forward and backward) versus ear and sequential OoR strategies (right ear, left ear, sequential, and ear or sequential). This was done by determining if the median difference in the occasions of use of temporal versus ear and sequential OoR strategies within participants differed from zero (note: the OoR strategies were split into temporal versus ear and sequential for ease of analysis and after suggestions from DD research in adults that channels [ears] were irrelevant in an auditory system and some ear OoR was in fact sequential OoR [Parkinson et al. 1974; Rollins, Schurman, and Everson 1972; Savin 1967]). The second set of analyses used Spearman’s rank order correlation coefficients at the 1% level (to mitigate the use of multiple comparisons) to identify correlations between 2pDD test scores and use of OoR strategy.

Non-parametric statistics were used for all inferential analyses to counter the presence of skewed data that could not be resolved by transformation with negative skews observed for six of the nine 2pDD test score variables and positive skews observed for participant age and five of the seven OoR strategies. All inferential analyses were conducted using IBM SPSS Statistics, version 28.0.0.0 (190) (IBM Corp 2021).

Ethics

Ethical clearance to conduct this clinical note was given by the Human Research Ethics Committee of The University of Queensland (approval number 2020000337).

Results

Table 1 shows the descriptive results for all participating children on each 2pDD test score. Figure 1 shows the proportional use of temporal, ear, sequential, and ear or sequential OoR for each participant. Individually, participants used these four OoR strategies in varying proportions ranging from the dominant use of one to the varied use of all four strategies. As a group, the participants used temporal OoR strategies more than ear and sequential OoR strategies (T = 1746.5, z = 2.18, p = 0.029, r = 0.25 [a small effect size]). Within the temporal OoR strategies only, participants used temporal forward more than temporal backward OoR (T = 2489.5, z = 6.95, p < 0.001, r = 0.79 [a large effect size]). Within the ear OoR strategies only, participants used right ear more than left ear OoR (T = 1269.5, z = 3.56, p < 0.001, r = 0.41 [a moderate effect size]). Note effect sizes were categorised using the arbitrary scale of 0.10 to <0.30 = small effect, 0.30 to <0.5 = moderate effect, and ≥0.5 = large effect.

Figure 1. Proportions of OoR strategies used by each participant. Participants returning total proportions <100% included response types not able to be classified in this clinical note. Temporal strategies were used more than ear and sequential strategies (T = 1746.5, z = 2.18, p = 0.029, r = 0.25 [a small effect size]).

Table 1. Descriptive results for 2pDD test scores (n = 77, median age = 8.8 years, IQR for age = 3.4 years).

	Score (%)
Scored item	Median	1Q	3Q	Minimum	Maximum
Digits in the test	86.25	80.0	93.75	53.75	100
All digits in a test item	50	30	75	0	100
Both digits to RE in a test item	92.5	87.5	97.5	57.5	100
Both digits to LE in a test item	82.5	72.5	92.5	50	100
Ear advantage (RE – LE)	7.5	0	15	−17.5	45
1^st digit to RE in a test item	95	90	100	55	100
2^nd digit to RE in a test item	90	85	100	50	100
1^st digit to LE in a test item	90	75	95	45	100
2^nd digit to LE in a test item	75	65	90	45	100

LE = left ear, RE = right ear.

Table 2 shows the Spearman’s rank order correlation co-efficients for age and 2pDD test scores versus age and use of OoR. In general, 2pDD test scores were positively correlated with age and negatively correlated with use of ear OoR.

Table 2. Spearman’s rank order correlation co-efficients for age and 2pDD test scores versus age and use of OoR strategy (n = 77).

Age and type of correct response	Age	Use of OoR strategy
		Temporal forward	Temporal backward	Right ear	Left ear	Sequential	Ear or sequential
Age		.10	−.12	−.16	−.17	.13	.13
Digits in the test	.34*	.23	−.07	−.58*	−.27	−.16	.19
All digits in a test item	.36*	.22	.04	−.62*	−.24	−.18	.14
Both digits to RE in a test item	.36*	.24	−.16	−.30*	−.43*	−.15	.19
Both digits to LE in a test item	.28	.20	−.02	−.61*	−.10	−.18	.10
Ear advantage (RE - LE)	−.03	−.04	−.12	.45*	−.17	.11	.12
1^st digit to RE in a test item	.34*	.22	−.24	−.35*	−.38*	−.19	.26
2^nd digit to RE in a test item	.35*	.20	−.02	−.23	−.40*	−.07	.13
1^st digit to LE in a test item	.19	.43*	.01	−.73*	−.11	−.23	−.04
2^nd digit to LE in a test item	.33*	−.01	−.06	−.34*	−.08	−.09	.21

*p ≤ .01, RE = right ear, LE = left ear.

Discussion

When completing 2pDD testing in the free recall condition, a large sample (n = 77) of children aged 7.0 to 17.4 years who had been referred for an AP assessment showed: 1) greater use of temporal OoR (particularly temporal forward) than ear or sequential OoR, and 2) decreased 2pDD test scores with increased use of ear OoR strategies (particularly right ear).

Whilst speculative, the participating children could have favoured the temporal forward OoR strategy (where digits from the first pair were reported before digits from the second pair) for the opportunity it provides to rehearse the first digit pair (possibly strengthening its memory trace) prior to processing the second digit pair. The significance of this group preference was mitigated by individual children using varying combinations of OoR strategies ranging from the almost exclusive use of a single OoR strategy to the use of almost all OoR strategies within the single test session. Such variation was consistent with previous reports of children using multiple OoR strategies during DD test sessions with choice of strategy possibly the result of individual preference (Neufeldt 1966; Surwillo et al. 1972; Witelson and Rabinovitch 1971).

The decreased 2pDD test scores with increased use of ear OoR strategies (particularly right ear) were more difficult to explain. One postulation is the use of ear OoR could exacerbate the ear dominance effects expected in dichotic processing especially in younger children (Moncrieff 2011), although no correlation was noted between ear advantage and age in the present note’s participating children. This postulation is partially supported by two findings. First, the right ear OoR returned its largest negative correlations with left ear test scores and its only positive correlation with the ear advantage score (calculated as right ear score minus left ear score). Second, the left ear OoR only returned negative correlations with right ear test scores.

Finally, it was interesting to note that participant age correlated with multiple 2pDD test scores but not with any OoR strategy. The positive correlations between age and 2pDD test scores were consistent with the known improvement in dichotic listening with age (with adult-like performance levels expected in healthy children by the pre- to early teenage years (Cameron and Dillon 2020; Cameron, Glyde, Dillon, and Whitfield 2016; Moncrieff 2011; Musiek 1983; Weihing et al. 2015; Westerhausen and Hugdahl 2008). The lack of correlation between age and OoR strategies suggests OoR preferences may not change substantially with age in clinical samples of children referred to audiology for an AP assessment.

The present clinical note contained both strengths and limitations. The convenient sampling of the participants from a clinical population was both a strength in that it allowed the assessment of a population of children likely to undergo 2pDD testing and a limitation in that it exposed the study to sampling bias. The lack of consideration of the handedness of the participating children was also a limitation in that it prevented any checks for possible effects of handedness on patterns of laterality in the 2pDD test results (Moncrieff and Musiek 2002).

Conclusions

When completing a 2pDD test in the free recall condition, a large sample of children referred to audiology for an AP assessment used temporal more than ear or sequential OoR and showed decreased test scores with increased use of ear OoR. Whilst postulative, these results suggest these children could have favoured a temporal OoR for its opportunity to rehearse the first digit pair (possibly strengthening its memory trace) prior to processing the second digit pair, while their use of an ear OoR could have come at the cost detrimental ear dominance effects that reduced their test performance. Future research could consider if similar results are observed in other child populations and if preferred dichotic processing strategy could inform auditory training for children with dichotic listening deficits.

Author contributions

Wayne J. Wilson and Sarosh Kapadia: Conceptualisation (Equal), Data curation (Equal), Formal analysis (Equal), Funding acquisition (Equal), Methodology (Equal), Project administration (Equal), Resources (Equal), Software (Equal), Supervision (Equal), Validation (Equal), Visualisation, Writing – original draft (Equal), and Writing – review & editing (Equal). Auriel Sher and Roisin Higgins: Data curation (Equal), Investigation (Equal), Project administration (Equal), Resources (Equal), Software (Equal), Validation (Equal), and Writing – original draft (Equal).

Acknowledgements

We thank Samuel Baird, Hee Gyeong Kim, Hephzibah Thankam Koothoor Simon, Samuel Madigan, Xiankai Meng and Joseph Reeks, who worked on parts of this dataset for their second-year Master of Audiology research projects at Flinders University in Adelaide, Australia.

Data access statement

The datasets generated during and/or analysed during the current clinical note are available from the corresponding author on reasonable request.

Disclosure statement

No potential conflict of interest was reported by the authors.

Additional information

Funding

This research was provided with in-kind support by The University of Queensland and Flinders University, Australia.