International Journal of Pediatric Otorhinolaryngology 110 (2018) 93–99

Contents lists available at ScienceDirect

International Journal of Pediatric Otorhinolaryngology

journal homepage: www.elsevier.com/locate/ijporl

The Persian version of auditory word discrimination test (P-AWDT) for T

children: Development, validity, and reliability
Nassim Hashemia, Ali Ghorbanib, Zahra Soleymania, Mohmmad Kamalic,
Zohreh Ziatabar Ahmadid,∗, Saeid Mahmoudiane,f
a
Department of Speech Therapy, School of Rehabilitation, Tehran University of Medical Sciences, Tehran, Iran
b
Department of Speech Therapy, School of Rehabilitation Sciences, Iran University of Medical Sciences, Tehran, Iran
c
Department of Basic Sciences in Rehabilitation, School of Rehabilitation Sciences, Iran University of Medical Sciences, Tehran, Iran
d
Department of Speech Therapy, School of Rehabilitation, Babol University of Medical Sciences, Babol, I.R, Iran
e
Laboratory for Auditory Neuroscience, ENT and Head & Neck Research Center, Iran University of Medical Sciences, Tehran, Iran
f
Department of Otorhinolaryngology - Medical University of Hannover (MHH), Hannover, Germany

A R T I C LE I N FO A B S T R A C T

Keywords: Objective: Auditory discrimination of speech sounds is an important perceptual ability and a precursor to the
Auditory discrimination acquisition of language. Auditory information is at least partially necessary for the acquisition and organization
Test of phonological rules. There are few standardized behavioral tests to evaluate phonemic distinctive features in
Validity children with or without speech and language disorders. The main objective of the present study was the de-
Reliability
velopment, validity, and reliability of the Persian version of auditory word discrimination test (P-AWDT) for 4-8-
Speech sound disorder
year-old children.
Methods: A total of 120 typical children and 40 children with speech sound disorder (SSD) participated in the
present study. The test comprised of 160 monosyllabic paired-words distributed in the Forms A-1 and the Form
A-2 for the initial consonants (80 words) and the Forms B-1 and the Form B-2 for the final consonants (80
words). Moreover, the discrimination of vowels was randomly included in all forms. Content validity was cal-
culated and 50 children repeated the test twice with two weeks of interval (test–retest reliability). Further
analysis was also implemented including validity, intraclass correlation coefficient (ICC), Cronbach's alpha
(internal consistency), age groups, and gender.
Results: The content validity index (CVI) and the test-retest reliability of the P-AWDT were achieved 63%–86%
and 81%-96%, respectively. Moreover, the total Cronbach’s alpha for the internal consistency was estimated
relatively high (0.93). Comparison of the mean scores of the P-AWDT in the typical children and the children
with SSD revealed a significant difference. The results revealed that the group with SSD had greater severity of
deficit than the typical group in auditory word discrimination. In addition, the difference between the age groups
was statistically significant, especially in 4–4.11-year-old children. The performance of the two gender groups
was relatively same.
Conclusion: The comparison of the P-AWDT scores between the typical children and the children with SSD
demonstrated differences in the capabilities of auditory phonological discrimination in both initial and final
positions. It supposed that the P-AWDT meets the appropriate validity and reliability criteria. The P-AWDT test
can be utilized to measure the distinctive features of phonemes, the auditory discrimination of initial and final
consonants and middle vowels of words in 4-8-year-old typical children and children with SSD.

1. Introduction redundancy sounds” [1]. Auditory discrimination is commonly referred

The processes of the auditory discrimination include “location and
lateralization of sound, recognition of auditory patterns, temporal as-
pects of hearing (temporal integration, temporal discrimination, tem-
poral order, and marking), and perception of competitive and low-
to as the capacity to distinguish between phonemes or individual
speech sounds [2]. It is emphasized that the precise dissociation of
spoken sounds is the critical factor for the achievement of the articu-
latory proficiency and the language comprehension [3–5]. Six-month-
old infants are able to detect consonant contrasts and able to

∗
Corresponding author. Department of Speech Therapy, School of Rehabilitation, Babol University of Medical Sciences, Ganj Afrooz Ave, Postal Code: 47745-47176, Babol, Iran.
E-mail addresses: z.ziatabar@mubabol.ac.ir, ziatabar.ahmadi@gmail.com (Z.Z. Ahmadi).

https://doi.org/10.1016/j.ijporl.2018.05.003
Received 19 January 2018; Received in revised form 30 April 2018; Accepted 3 May 2018
Available online 05 May 2018
0165-5876/ © 2018 Elsevier B.V. All rights reserved.
N. Hashemi et al.
differentiate among the phonemes of language which they are exposed
[6–9]. According to the Maintenance theory, the auditory experience is
required to maintain perceptual sensitivities in the infants [10].
Therefore, the discrimination ability may be improved through the
language experiences, the developmental changes, and age [6–9]. It has
been hypothesized to be a hierarchical and maturational pattern of the
phonetic distinction in infants and children [11]. For example, the re-
searchers revealed that manner and voicing is the most important au-
ditory dimensions which followed by the place of articulation [12].
Young children can separate the acoustic components of the manner
and the voicing better than the acoustic components of place of ar-
ticulation [13]. Moreover, it can be inferred that vowel changes are
occurred earlier than the consonant changes [11].
In order to acquire the speech sounds, children are needed to have a
typical oral, auditory, and central nervous systems [3]. A speech dis-
order is characterized by inadequate use of the phonological rules of
language (DSM-IV F80.0–315.39) [14,15]. Speech sound disorders
(SSD) may be accompanied by problems in the process of the phonemic
differentiation and comprehension [3,16,17]. Children with SSD have
difficulties to produce the speech sounds and the phonological re-
presentations, and to organize the lingual rules that are characterized
by errors such as phoneme substitution, omission, and distortion
[3,16,17]. According to the theories concerning the speech sound
production and comprehension, the SSD can be generated by the central
auditory processing problems, difficulties in the acquirement of the
phonological rules and the speech production [14,18]. Accordingly, the
precise observation of the speech performance in children with SSD can
partially contribute in understanding of their speech and language
abilities. Although the relationship between speech-sound discrimina-
tion and the SSD are still unknown and controversial [19,20], it sus-
pected that there is a reciprocal association between the articulation
disorder and the auditory discrimination, as confirmed by Liberman's
motor mediation theory [21]. According to the Liberman's theory [21],
“the phonetic information is perceived in a biologically distinct system,
a ‘module’ specialized to detect the intended gestures of the speaker
that are the basis for the phonetic categories”. As results, the lack of
valid and reliable data, especially concerning the capabilities of Per-
sian-speaking children with SSD in the distinction of the auditory seg-
mental units led to the need for a reliable instrument.
The instruments that are used to measure speech discrimination in
children have been relatively limited. The most widely used instru-
ments were the Goldman-Fristoe-Woodcock Test of Auditory
Discrimination [22], the Wepman Auditory Discrimination Test (ADT)
[2], and the Boston University Speech-Sound Discrimination Test [23].
The Wepman test includes a comparison of paired-words by auditory
presentation while the Goldman-Fristoe and Woodcock test contains a
language-context test [2]. This instrument utilizes a carrier phrase or
sentence context to present auditory stimuli and the child should select
a picture from a category [2]. In the Wepman's test, the child is asked to
response if two words presented by auditory method are the same or
different [2]. This test provides precise information regarding a child's
speech-sound discrimination and is more difficult than the Goldman-
Fristoe-Woodcock test. On the other hand, the ADT consists of two al-
ternative forms (Form A1 and Form A2) [2]. This test is a very rapid
screening measure for the identification of children with auditory dis-
crimination problems [2]. However, the ADT investigates distinctive
features of phonemes only in the initial position of words [2]. The
Boston University Speech-Sound Discrimination Test is a picture-word
discrimination test used for young children without considering dis-
tinctive features and specific phonological rules [23]. The tests of au-
ditory discrimination are generally old and often designed for English-
speaking children. Each language seems to have unique multifarious
phonological and metalinguistic rules. Therefore, it is logical that au-
ditory discrimination tests should be designed based on the lingual and
the cultural characteristics.
To the best of our knowledge, the understanding and the
94
International Journal of Pediatric Otorhinolaryngology 110 (2018) 93–99
implementation of speech sound discrimination tests are commonly
heavy for young children [24]. One possible reason is that the pure
auditory tests provide the confined lingual information concerning to
the phonological discrimination [24]. Moreover, auditory discrimina-
tion tests are different from each other by multiple factors. These fac-
tors contain the mode of presentation (auditory or visual), the position
of the phonemic contrasts (initial, middle, or final consonant), and the
type of distinctive features (placement, voicing, and manner), the type
of segment (consonants or vowels), syllabic context, age groups
[2,18,22,23]. On the other hand, a listener's correct identification of a
single spoken word depends on the situation, dialects, and the mode of
presentation by examiner. Therefore, providing of a similar mode of
presentation for all the participants, such as a recorded voice can ap-
proximately decrease biases in the research studies. Accordingly, it
seems a necessity to assess the skills of auditory phonological dis-
crimination using minimal sets in other languages.
The aim of the present study to develop a comprehensive list of
recorded and Persian words with different vowels, as well as initial and
final consonants based on the distinctive features of phonemes.
Furthermore, the various age groups (4–8 years) are examined to follow
the probable developmental trends of the auditory word discrimination
in the typical children and the children with SSD.
2. Materials and methods
2.1. Participants
A total of 120 typical children, aged between 4 and 8 years
(Mean = 5.62 ± 1.22), and 40 children with SSD, aged between 4 and
8 years (Mean = 5.57 ± 1.12), participated in the present study. All
the children were monolingual and native speakers of Persian language
and were selected from Tehran kindergarten and Primary schools.
Typical children were healthy, socially well-functioning, and none had
neither cognitive nor learning difficulties; children with SSD had only
speech sound errors as a disorder. The diagnosis was established in-
formally by a professional speech and language pathologist in a spe-
cialized clinic affiliated with the researchers' university based on client
articulation errors. The diagnostic criteria were based on (DSM-IV
F80.0–315.39) [14,15]. The criteria were persistent unintelligible
speech consisting of phoneme addition, omission, distortion, or sub-
stitution, which interferes with verbal communication. The two groups
were matched for age, gender, and parental socio-economic status.
Participants' characteristics are shown in Table 1. All children were
evaluated in a private room with the experimenter present. The as-
sessment for each participant took place in one session, as well.
2.2. Ethical considerations
The study was approved by the Ethical Committee of the Tehran
University of Medical Sciences. All parents enrolled in the study filled a
written informed consent which was designed according to the
Table 1
Characteristics of the participants with age means (and standard deviations).
Group Gender N Chronological Age
Mean(SD)
Typical F 64 5.68 (1.26)
M 56 5.55 (1.18)
Total 120 5.62 (1.22)
SSD F 20 5.60 (1.19)
M 20 5.53 (1.08)
Total 40 5.57 (1.12)
Total F 84 5.61 (1.25)
M 76 5.47 (1.16)
Total 160 5.61 (1.19)
N. Hashemi et al. International Journal of Pediatric Otorhinolaryngology 110 (2018) 93–99

Declaration of Helsinki.
2.3. Persian version of auditory word discrimination test (P-AWDT)
In order to administer the P-AWDT, we used from a list of paired-
words with consonant-vowel-consonant monosyllabic structures. The P-
AWDT inspired by Wepman's Auditory Discrimination Test [2]. The
meaningful word list was prepared from a collection of story books
appropriate for the children. The difference between paired-words was
in initial or final consonants and middle vowels. Initial and final con-
sonants were different from each other in minimal sets (the voicing, the
manner, and the place of articulation). Although meaningless pseudo-
words were substituted with meaningful words, they did not impact on
performance and children didn't make errors on these items. To eval-
uate ability in the distinction of different pairs and the diagnosis of the
same pairs, 25% of all the words were selected as same pairs that dis-
tributed randomly through the four forms. Overall, 160 paired-words in
four forms (Form A-1, Form A-2, Form B-1, and Form B-2) were selected
as the test items (e.g. 10 sample of paired-words from each category
listed in Appendix 1). Each form is contributed from 40 items. Forms A-
1 and Form A-2 were designed to survey the initial phonemes and
Forms B-1 and Form B-2 were developed to investigate the final pho-
nemes. The vowel changes (16 items) were also assessed and dis-
tributed randomly through the four forms. The total P-AWDT scores
were in the range of 0 and 160. A native Persian male speaker produced
the final list of 160 words several times at a 44.1 kHz sampling rate
with neutral intonation in a sound-proof studio. A microphone (Model:
M-Audio Nova) was used set at 20 cm distance from the speaker's
mouth. All the recorded words were transferred to a sound card (Model:
M-Audio Fast track pro II) and all the recordings were edited using
Behringer Mixer (Model: XeroX1002FX) and Cubase, Logic software.
The duration between the two words in one pair and the two paired-
words from each other were 500 ms and 3 s, respectively. The intensity
of sound was measured at the height of the listener's head using a Sound
Level Meter (Model: GM 1357-EN-00). Sound intensity level was set on
70 dB SPL. All the stimuli were presented a 10 inch laptop (Aspire one
3665-D257) in a pseudorandom sequence to control for order/learning
effects. Two high-quality headphones placed in the child's ear were
used to present. A primary list of items (10 items) was available for the
children to use prior to presenting main items. Children were asked to
verbally discriminate between each paired-word presented by re-
sponding “yes” inferred to mean “same”) or “no” to indicate “different”.
reliability of this test [26]. For test-retest reliability, 50 participants
were re-examined twice with two weeks of interval and the correlations
between the test-retest mean scores were compared. Reliable perfor-
mances on the tasks result in high scale correlations across the test-
retest interval. Intra-class correlation (ICC) higher than 70% was con-
sidered as acceptable [27].
2.6. Statistical analyses
Kolmogrolov-Smirnov was used to ensure normal distribution of the
samples. Furthermore, the CVI was calculated to examine the content
validity of the P-AWDT and the ICC and the Pearson's correlation
coefficient were used to check the test-retest reliability. Besides, de-
scriptive and analytic statistics were used. For descriptive analysis, the
mean scores and standard deviation values were used. Due to non-
normal distribution of samples, non-parametric Mann-Whitney U,
Kruskal-Wallis, and Chi square tests were used for the comparison of the
P-AWDT scores in the age groups and genders for the typical children
and the children with SSD. Data was analyzed using SPSS 19 (SPSS,
Inc., Chicago, IL) and the statistical significance was set at p ≤ 0.05.
3. Results
The results of the present study are presented in three parts. In first
and second parts, we reported the results of the content validity, the
test-retest reliability, and the internal consistency of the P-AWDT. In
third part, the results of the correlation between the P-AWDT and
gender were illustrated by a box plot. Finally, all children's perfor-
mances including the descriptive and analytic values on individual
tasks in the four age groups (4–4.11, 5–5.11, 6–6.11, 7-8 year-old) are
summarized in tables and a simple line chart.
3.1. Content validity
The CVI of each form is presented in Table 2. The CVIs for the Form
A-1 and the Form A-2 were achieved 63% and 85%, respectively, and
the CVIs for the Form B-1 and the Form B-2 were estimated 60% and
86%, respectively. The results demonstrated that the P-AWDT has
strong and acceptable content validity in all the four forms.

2.4. Validity of the P-AWDT 3.2. Internal consistency and test-retest reliability

Content validity of the P-AWDT was evaluated by a panel of 14 Total Cronbach's alpha of the P-AWDT was 0.93. Table 2 presents
judges. For each question, a 0-2-point scale was used; (2 points for: the means, standard deviations, the ICC, and Pearson product moment
“very suitable”, 1 for: “somewhat suitable”, and 0 for: “totally un- correlations between the test and retest with two weeks of interval in 50
suitable” [25].The criteria were used to judge the suitability of each typical children. The ICC measures were observed between 84% and
item based on the types of selected consonants and vowels. Then, we 96% in the four forms. Furthermore, there was a significant correlation
computed the content validity index (CVI) for each item: only items between children's performances in first and second times. These data
scoring 2 were regarded as acceptable. Therefore, the CVI was calcu- reflected the Pearson product-moment correlations in all the forms
lated as following: the number of judges who scored 2 divided by the which were found very high (0.81 ≤ r ≥ 0.96).
total number of judges. Only items over 0.51 in the CVI, meaning that
Table 2
more than 51% of the experts regarded them as valid, were selected as
The results of the content validity and the test-retest reliability of the P-AWDT.
significant [25]. Other items with low score were eliminated.
P-AWDT Validity Test Retest Reliability
2.5. Reliability of the P-AWDT
CVI M SD M SD ICC Correlation
Coefficients
Reliability of the P-AWDT was determined in two ways: Internal
consistency and test-retest reliability. To evaluate the internal con- Form A-1 63% 38.85 1.14 39.02 0.99 94% 0.96
sistency, we examined inter-task agreement by Cronbach alpha. In this Form A-2 85% 39.25 0.80 39.25 0.77 96% 0.96
Form B-1 60% 39.12 1.01 39.12 0.96 84% 0.81
procedure, each task was essentially treated as a test item. When the Form B-2 86% 39.20 0.91 39.37 0.86 84% 0.82
different tasks reliably measure the same construct, Cronbach's alpha
will be high. When each word is individually deleted, alpha never drops CVI = Content validity index; M = mean; SD = Standard deviation, ICC =
below 0.70, suggesting that all the words are contributing equally to the Intraclass correlation.

95
N. Hashemi et al. International Journal of Pediatric Otorhinolaryngology 110 (2018) 93–99

Table 4
The percentage of correct responses to vowels changes compared with con-
sonants in the typical children and the children with SSD.
P-AWDT Mean Rank Mann-Whitney Wilcoxon Z P-value

Typical SSD

Form A-1 90.60 50.14 1185.50 2005.50 −5.04 0.000*

Form A-2 89.63 53.13 1305.00 2125.00 −4.67 0.000*
Form B-1 91.66 47.03 1061.00 1881.00 −5.68 0.000*
Form B-2 91.00 48.99 1139.50 1139.50 −5.37 0.000*

Note. *P ≤ 0.05.

Fig. 1. The mean ranks of the scores of auditory word discrimination skills
(Form A-1, Form A-2, Form B-1, and Form B-2) are reported in all the parti-
cipants in the two genders by the line chart. There was no significant difference
in the mean scores of the P-AWDT between the two gender groups.
3.3. Gender and the P-AWDT
AWDT in the four age groups (4–4.11, 5–5.11, 6–6.11, and 7–8 years)
and in the typical children and in the children with SSD by a simple line
chart. The SEMs in all the age groups were larger in the children with
SSD than in the typical children. The mean scores of each age group
showed that the capabilities of auditory word discrimination are line-
arly increased when children grow. This indicates that the P-AWDT has
the essential criteria of a developmental scale and is sensitive to ma-
turation. As can be seen by chart, the most increase in the auditory
word discrimination took place between the ages of 6 and 8 years.

The comparison of gender differences in the P-AWDT using non-

3.5. The comparison of the vowels and the consonants
parametric Mann-Whitney U test were obtained for Form A-1
(z = −0.16; p = 0.86), Form A-2 (z = −0.12; p = 0.90), Form B-1
Table 5 presented the percentage of correct responses to the vowel
(z = −0.28; p = 0.77), and Form B-2 (z = −0.21 (p = 0.82). Besides,
changes which were compared with consonants in the typical children
Fig. 1 reflected the description of the mean ranks of the P-AWDT in all
and the children with SSD using Chi square test. The typical children
the girls and the boys in the four forms by a line chart. Overall, the
and the children with SSD were answered to the vowel changes better
comparisons between the boys and the girls on each form revealed any
than the consonant discriminations in all forms (p < 0.05).
significant differences (p > 0.05).

3.4. Age and the P-AWDT 4. Discussion

As expected, there was a robust correlation between the P-AWDT
and age (Form A-1: r = 0.60, Form A-2: r = 0.52, Form B-1: r = 0.54,
and Form B-2: r = 0.52, p < 0.05). Tables 3 and 4 present the mean
scores, standard deviations, and the values of P-value in the four age
groups in typical children and those with SSD by nonparametric Mann-
Whitney U, Kruskal-Wallis, and Chi square tests. The correlation be-
tween the P-AWDT and age was most commonly significant in the
4–4.11-year-old group (p = 0.00). The difference between the typical
children and the group with SSD was statistically significant in all the
four forms (p = 0.00). Furthermore, Fig. 2 shows the comparison of the
mean scores and the standard errors of measurements (SEM) of the P-
Undoubtedly, the distinction of speech-related auditory stimuli as
one of neurobiological pathways has directly been implied in auditory
processing [28–30]. During the recent years, it has been a necessity to
consider the probable reasons, basic generators, and all the con-
sequences of a speech and/or language disorder. The present study was
carried out in order to develop a Persian version of auditory word
discrimination test to measure phonemic distinctive features in typical
children and children with SSD.
The most important characteristics of a valuable tool are repeat-
ability, consistency, and validity of obtained scores [25,31]. In in-
vestigation of content validity, Polit et al. [25] and Shi and et al. [32]

Table 3
The comparison of the means scores of the subtests in all age groups in the typical children and the children with SSD.
P-AWDT Age Typical SSD Chi-scores Mean Rank P-value

Mean (SD) Mean (SD) 4–4.11 5–5.11 6–6.11 7–8

∗
Form A-1 4–4.11 36.90 (2.02) 33.40 (3.09) 58.04 35.45 1.00 0.00 0.00* 0.00*
5–5.11 39.40 (0.89) 37.70 (0.94) 87.43 0.00* 1.00 0.70 0.74
6–6.11 39.66 (0.54) 38.50(1.50) 100.03 0.00* 0.70 1.00 1.00
7–8 39.66 (0.54) 38.40 (1.26) 90.10 0.00* 0.74 1.00 1.00
Form A-2 4–4.11 37.90 (1.49) 34.50 (2.95) 63.44 35.86 1.00 0.00* 0.00* 0.00*
5–5.11 39.56 (0.67) 38.10 (1.37) 84.15 0.00* 1.00 0.25 0.62
6–6.11 39.86 (0.34) 38.10 (1.37) 106.00 0.00* 0.25 1.00 0.91
7–8 39.76 (0.43) 39.00 (0.81) 95.99 0.00* 0.62 0.91 1.00
Form B-1 4–4.11 37.56 (1.45) 34.90 (2.02) 60.60 35.46 1.00 0.00* 0.00* 0.00*
5–5.11 39.60 (0.72) 37.80 (1.22) 87.46 0.00* 1.00 0.56 0.50
6–6.11 39.76 (0.43) 38.80 (1.31) 98.58 0.00* 0.56 1.00 1.00
7–8 39.86 (0.34) 38.60 (1.07) 100.50 0.00* 0.50 1.00 1.00
Form B-2 4–4.11 37.86 (1.54) 35.10 (1.37) 55.67 37.40 1.00 0.00* 0.00* 0.00*
5–5.11 39.60 (0.62) 38.40 (1.07) 87.88 0.00* 1.00 0.74 0.88
6–6.11 39.83 (0.37) 38.80 (1.03) 99.49 0.00* 0.74 1.00 0.99
7–8 39.80 (0.40) 38.60 (1.26) 97.24 0.00* 0.88 0.99 1.00

Note. *P ≤ 0.05.

96
N. Hashemi et al. International Journal of Pediatric Otorhinolaryngology 110 (2018) 93–99

lower in the Form A-1 and the Form B-1 than in the Form A-2 and the
Form B-2. These differences may be probably due to the type of selected
words and the opinion of experts. Our results suggested that the test-
retest reliability of the P-AWDT was relatively very high (≥0. 84). Like
the results of the current study, Wepman [2] reported a test-retest re-
liability of the ADT (0.91) for 4-8 year-old normal children. Besides, a
test-retest reliability of the ADT (0.88) with one to two weeks of in-
terval in children with ‘mild intellectual impairments’ was reported
[33]. Moreover, eventually, it became clear that the internal con-
sistency of the P-AWDT was very high (0.93). Internal consistency of
Goldman, Fristoe, and Woodcock's test [25] for Quit subtest was re-
ported between 0.51 and 0.88 and for Noise subtest between 0.63 and
0.68. Furthermore, Tafiadis et al. [34] demonstrated the reliability of
the Greek minimal pairs test by Cronbach's alpha which estimated
72.16%. It implies that the response of every item to the response of the
total items in the P-AWDT was matched. These findings suggest that the
P-AWDT is a reliable instrument of auditory discrimination for typical
developing children and children with SSD. In other hands, it can be
concluded that the instrument can have satisfactory application as part
of a clinical or education test battery in clinical/research environments.
In addition, the developmental difference, which observed in the
children's auditory discrimination ability, suggests a maturational and
developmental shift effect. This difference was especially displayed in
the 4–4.11-year-old age group. The overall magnitude of effect in other
groups was small. The performance of this group compared to other
groups was low in the discrimination of initial and final consonants and
vowel changes. The discrimination of vowel changes was partially
better than the consonant discrimination in all groups and all four
forms. It is probably due to the different acoustic cues which distinguish
these two classes of speech sounds. As a result, the findings of the
current study were in accordance with those reported by Wepman [2]
who showed a minimal effect of age on the auditory discrimination
skills. It is well recognized that the young children perform more poorly
than older children and adults on psychoacoustic tasks. Their poor
performance could to be due to both sensory and non-sensory compo-
nents [35].
Yet further analysis provides important insights into the relationship
between the gender differences on the P-AWDT scores. Like Wepman's
[2] findings, the performances of both genders on the P-AWDT were
same. Moreover, the evidence of the current study suggested that the
typically developing children obtained the maximal results, especially
Fig. 2. The mean scores of the P-AWDT in the four age groups (4–4.11, 5–5.11,
6–6.11, and 7–8 years) in the typical children (in the top) and the children with
in the initial and final consonants discrimination test. The children with
SSD (in the bottom). The SEMs are shown by the error bars. The mean scores of SSD were often encountered with problems in auditory analyses and
the P-AWDT are significantly lower in 4–4.11 year-old groups than in other age phonological representations and were completely unable to separate
groups in both the typical children and the children with SSD. The SEMs were the words with different phonemes from each other in while typical
larger in the children with SSD than in the typical children in all the age groups. developing children could carry out without problem. Consequently,
the participants have difficulties in retrieving the phonological re-
Table 5 presentations by the auditory feedback during speech production. We
The comparison of the mean ranks of the P-AWDT using nonparametric tests in found that the P-AWDT is able to discriminate between the typical
the typical children and the children with SSD. children and the children with speech articulation deficient. These re-
sults were homologous with those reported by Marquardt and Saxman
P-AWDT Typical SSD
[36] on the ADT.
Vowels Consonants P-value Vowels Consonants P-value Overall, it should be mentioned that problem in the acquisition of
phonological and auditory discrimination may result in an inability to
Form A-1 98.75 92.01 0.000* 94.37 87.35 0.000* decode and to organize the auditory stimuli. Due to the ambiguous
Form A-2 99.37 96.00 0.000* 100.00 94.00 0.000*
Form B-1 98.12 95.38 0.000* 97.50 93.40 0.000*
nature of neurophysiologic bases related to speech-sound discrimina-
Form B-2 99.16 97.05 0.000* 97.50 93.95 0.000* tion and numerous questions posed against the critical role of the dis-
crimination in the articulation disorders, the evaluations of phonemic
Note. *P ≤ 0.05. discrimination should be carried out in the typical developing children
and children with speech disorders for more accurate diagnosis and
advocated that the CVI is a simple validity method. Content validity of early intervention. The P-AWDT, as a valid and reliable auditory dis-
the P-AWDT was assessed acceptable. Unlike our study, the content crimination test, can be used to measure fine phonological differences
validity was not clearly demonstrated in many previous studies such as in 4-8-year-old Persian-speaking children. However, the researchers
Goldman, Fristoe and Woodcock's test [22]. Perhaps, the reason is that should be informed about the wide variety of the procedures provided
the most used words were selected on the basis of high familiarity and by the available auditory discrimination tests to select the relevant tool
meaningfulness for young children. However, the CVI was partially instrument according to the clinical and research requirements and

97
N. Hashemi et al.
applications.
In the current study, there were some limitations that should be
taken into account when popularize the findings. Auditory dis-
crimination skills should be considered in other population, as well.
Some factors could affect on the speech discrimination scores including
educational level, fatigue, confusion, self-confidence, linguistic cap-
abilities, and attention, etc. Furthermore, the examiner's ability to
correctly score participants' responses could have influenced the results.
Although the P-AWDT can be used as a valid and reliable test for
Persian-speaking typical children and children with SSD, the standar-
dization of the test should be carried out on larger sample sizes and on
other speech and language disorders. Finally, it is substantial to in-
vestigate other types of the validity and the reliability including the
construct and the concurrent validity, and the inter-rater reliability.
5. Conclusions
The inexpensive, confidential and brief nature of dimensional rating
scales makes these instruments attractive tools for research [37]. The
findings of the present research have shown that the P-AWDT appears
to be acceptable for validity and reliability. Moreover, a strong re-
lationship was observed, especially between the mean scores of 4–4.11-
year-old group and the P-AWDT. It suggests that there is a develop-
mental change in the acquisition of phonological distinctive features
Appendix 1
The samples of paired-words in each form are listed in the appendix 1.
Form A-1 Form A-2
čay-jay čin-jin
tar-kar tam-kam
gel-gul kel-kul
gom-gom gom-gom
sir-zir sud-zud
gaz-qaz gom-qom
niš-niš niš-niš
niz-riz niš-riš
kæm-xæm kam-xam
baz-boz bar-bor
References
[1] H.F. Wertzner, G.L. Claudino, D.E.S. Galea, L.K. Patah, M.M. Castro, Medidas
fonológicas em crianc¸as com transtorno fonológico, Rev. Soc. Bras. Fonoaudiol. 17
(2012) 189–195.
[2] J. Wepman, Auditory discrimination, speech, and reading, Elem. Sch. J. 60 (1960)
325–333.
[3] T.M. Attoni, V.G. Quintas, H.B. Mota, Evaluation of auditory processing and pho-
nemic discrimination in children with normal and disordered phonological devel-
opment, ORL (Oto-Rhino-Laryngol.) (Basel) 76 (2010) 762–768.
[4] C.N. Rocha-Muniz, E.C. Zachi, R.A. Teixeira, D.F. Ventura, D.M. Befi-Lopes,
E. Schochat, .Association between language development and auditory processing
disorders, Braz. J. Otorhinolaryngol. 80 (2014) 231–236.
[5] C. Cantiania, V. Rivaa, C. Piazzab, R. Bettonia, M. Moltenia, N. Choudhuryd,
C. Marinoa, A. Benasichd, Auditory discrimination predicts linguistic outcome in
Italian infants with and without familial risk for language learning impairment,
Dev. Cogn. Neurosci. 20 (2016) 23–34.
[6] P.K. Kuhl, Speech perception in early infancy: perceptual constancy for spectrally
dissimilar vowel categories, J. Acoust. Soc. Am. 66 (1979) 1668–1679.
[7] P.K. Kuhl, Early language acquisition: cracking the speech code, Nature. Rev.
Neurosci. 5 (2004) 831–843.
[8] D.P. Philips, Central auditory processing: a view from auditory neuroscience, Am. J.
Otolaryngol. 16 (1995) 338–352.
[9] H.F. Wertzner, L. Amaro, D.E.S. Galea, phonological performance measured by
speech severity indices compared with correlated factors, Sao. Paulo. Med. J. 125
(2007) 309–314.
[10] J.L. Anderson, J.L. Morgan, K.S. White, A statistical basis for speech sound dis-
crimination, Lang. Speech 46 (2003) 155–182.
98
International Journal of Pediatric Otorhinolaryngology 110 (2018) 93–99
with age. This test can make a distinction between normal children and
children with SSD. The children with SSD in comparison to the typical
children had greater difficulty in cognitive-linguistic processing (orga-
nization of phonological rules), auditory processing, and/or speech
production. There was no significant correlation between auditory
discrimination and gender. The test can be practically used to measure
and to diagnosis the phonological auditory discrimination difficulties in
Persian-speaking and typical children, children with SSD, and probably
other children with various auditory processing disorders.
Funding
This research received no specific grant from any funding agency in
the public, commercial, or not-for-profit sectors.
Conflicts of interest
The authors declare that there is no conflict of interests.
Acknowledgment
We would like to express our sincere thanks to all the parents and
children who have participated in the present study.
Form B-1 Form B-2
kot-kod sut-sud
baz-bar suz-sur
gel-gol qel-qol
šæb-šæb šæb-šæb
gav-gaz nav-naz
sæq-sæg tæq-tæg
tæk-tæk tæk-tæk
sæd-sag ræd-ræg
maš-mah kaš-kah
tær-tir dær-dir
[11] P.K. Kuhl, Brain mechanisms in early language acquisition, Neuron 67 (2010)
713–727.
[12] N. Bedoin, E. Ferragne, E. Marsico, Hemispheric asymmetries depend on the pho-
netic feature: a dichotic study of place of articulation and voicing in French stops,
Brain Lang. 115 (2010) 133–140.
[13] S.Z. Ziatabar Ahmadi, S. Mahmoudian, H. Ashayeri, F. Allaeddini, M. Farhadi,
Electrophysiological and phonological change detection measures of auditory word
processing in normal Persian-speaking children, Int. J. Pediatr. Otorhinolaryngol.
90 (2016) 220–226.
[14] T.F. Barrozo, L.O. Pagan-Neves, N. Vilela, R.M. Carvallo, H.F. Wertzner, The in-
fluence of (central) auditory processing disorder in speech sound disorders, Braz, J.
Oto-Rhino-Laryngol. 82 (2016) 56–64.
[15] N. Vilela, T.F. Barrozo, L.O. Pagan-Neves, S.G. Sanches, H.F. Wertzner,
R.M. Carvallo, The influence of (central) auditory processing disorder on the se-
verity of speech-sound disorders in children, Clinics 72 (2016) 62–68.
[16] M. Benjamin, M.O.P. Krause, Phonological encoding in speech sound disorder:
evidence from a cross-modal priming experiment, Int. J. Lang. Commun. Disord. 52
(2017) 285–300.
[17] S. Hearnshaw, E. Baker, N. Munro, The speech perception skills of children with and
without speech sound disorder, J. Commun. Disord. 71 (2017) 61–71.
[18] N.G. Bountress, C.M. Laderberg, A comprison of two tests of speech sound dis-
crimination, J. Commun. Disord. 14 (1981) 149–156.
[19] S. Masso, S. McLeod, E. Baker, J. McCormack, Polysyllable productions in preschool
children with speech sound disorders: error categories and the framework of
polysyllable maturity, Int. J. Speech Lang. Pathol. 18 (2016) 272–287.
[20] T.P. Marquardt, J.H. Saxman, Language comprehention and auditory discrimina-
tion in articulation deficient kindergarten children, J. Speech Lang. Hear. Res. 15
(1972) 382–389.
[21] A.M. Liberman, I.G. Mattingly, The motor theory of speech perception revised,
N. Hashemi et al.
Cognition 21 (1985) 1–36.
[22] R. Goldman, M. Fristoe, R.W. Woodcock, Goldman-Fristoe-Woodcock test of audi-
tory discrimination. american guidance service, J. Special. Educ. 4 (1973) 367–373.
[23] W. Pronovost, The Boston University Speech Sound Discrimination Test, (1974)
Cedar Falls IA: Go-Mo Products.
[24] M.A. Geissal, J.D. Knafle, A linguistic view of auditory discrimination tests and
exercises, Read. Teach. 31 (1997) 134–141.
[25] D.F. Polit, C.T. Beck, S.V. Owen, Is the CVI an acceptable indicator of content va-
lidity? Appraisal and recommendations, Res. Nurs. Health 30 (2007) 459–467.
[26] J. Reynaldo, A. Santos, Cronbach's alpha: a tool for assessing the reliability of scales,
J. Extention 37 (1999) 1–5.
[27] M. Yen, L.H. Lo, Examining test-retest reliability: an intra-class correlation ap-
proach, Clin. Nurs. Res. 51 (2002) 59–62.
[28] G. Hickok, D. Poeppel, Towards a functional neuroanatomy of speech perception,
Trends. Cogn. Sci. 4 (2000) 131–138.
[29] S.K. Scott, How might we conceptualize speech perception? The view from neu-
robiology, J. Phon. 31 (2003) 417–422.
[30] L.F. Halliday, O. Tuomainen, S. Rosen, Auditory processing deficits are sometimes
necessary and sometimes sufficient for language difficulties in children: evidence
from mild to moderate sensorineural hearing loss, Cognition 166 (2017) 139–151.
99
International Journal of Pediatric Otorhinolaryngology 110 (2018) 93–99
[31] N. Golafshani, Understanding reliability and validity in qualitative research, Qual.
Rep. 8 (2003) 597–606.
[32] J. Shi, X. Mo, Z. Sun, Content validity index in scale development, Zhong Nan Da
Xue Xue Bao Yi Xue Ban 37 (2012) 152–155.
[33] A.J. Dahle, D.A. Daly, Influence of verbal feedback on auditory discrimination test
performance of mentally retarded children, Am. J. Ment. Retard. 76 (1972)
586–590.
[34] D. Tafiadis, E. Chorozopoulou, E. Tsanousa, M. Tafiadi, The creation and pilot va-
lidationof the Greek test for auditory discrimination (in minimal paires) for pre-
school children, Ann. Gen. Psychiatr. 7 (2008) 199.
[35] D.R. Moore, J.F. Rosenberg, J.S. Coleman, Discrimination training of phonemic
contrasts enhances phonological processing in mainstream school children, Brain
Lang. 94 (2005) 72–85.
[36] T.P. Marquardt, J.H. Saxman, Language comprehention and auditory discrimina-
tion in articulation deficient kindergarten children, J. Speech Lang. Hear. Res. 15
(1972) 382–389.
[37] M. Thastum, K. Ravn, S. Sommer, A. Trillingsgaard, Reliability, validity and nor-
mative data for the danish beck youth inventories, Scand. J. Psychol. 50 (2009)
47–54.