Speech-Language Pathologists’ Knowledge and Competence Regarding Percentage of Consonants Correct

Abstract

The aim of this article is to examine speech-language pathologists’ (SLPs) knowledge and competence using Percentage of Consonants Correct (PCC). Participants were recruited through posts on social media and message boards for the American Speech-Language-Hearing Association (ASHA) Special Interest Groups. In addition, surveys were emailed via the ASHA membership directory to SLPs in selected states who reported working in settings with a primarily pediatric population. The majority of SLPs reported a lack of academic and clinical training in the area of PCC. Participants demonstrated limited knowledge of the rules for calculating PCC in addition to decreased calculation ability and confidence. Because PCC is well-validated, SLPs would benefit from increased clinical and academic exposure to the measure; however, as many report that they do not use PCC, this calls into question its clinical practicality. The question remains if there is a need for a more expedient process by which to objectively quantify severity, such as use of an automated metric.

Keywords

speech-language pathologists assessment articulation speech/sound

Speech sound disorders (SSDs) are common in school-aged children. It is estimated that they are identified in up to 25% of kindergarten-age children (Tuomi & Ivanoff, 1977), and 90% of caseloads of school-based clinicians involve treatment of children with SSDs (American Speech-Language-Hearing Association [ASHA], 2018). The large percentage of children with SSDs indicates a need for knowledgeable clinicians who are competent in the assessment and treatment of this population. Most SLPs recognize that when determining eligibility for children with SSDs, many factors are a part of their “educational performance.” These factors include access to the curriculum; academic results/grades; oral participation in class; oral reading; spelling; social-emotional, adjustment/behavior; reaction of self, peers, teachers, or parents (Farquharson & Boldini, 2018). This finding highlights the understanding that SSDs can greatly impact a child’s education in a myriad of ways, both academically and socially.

ASHA (2004) has delineated a preferred practice pattern for the comprehensive assessment of SSDs that includes the use of standardized and nonstandardized measures such as articulation tests, spontaneous speech samples, analysis of error type, independent analysis (e.g., phonetic inventory) and relational analysis (e.g., phonological process analysis). Furthermore, ASHA suggests that examiners should also make a determination concerning severity. The severity of a child’s SSD is often judged on a continuum from mild to severe and may be assessed qualitatively or quantitatively. While ASHA advocates the use of a quantitative measure of severity, there is currently no consensus among professionals regarding the best way to assess severity of involvement (ASHA, n.d.; Flipsen, Hammer, & Yost, 2005).

While the standards of SSD assessment are delineated by ASHA (2004), there is little evidence as to whether these recommendations are currently being put into practice. In recent years, only two surveys investigated the practice patterns of speech-language pathologists (SLPs), as it pertains to the assessment of SSDs in English-speaking children (McLeod & Baker, 2014; Skahan, Watson, & Lof, 2007). The findings from these evaluations indicate that clinicians typically rely on parent interviews, speech samples, commercially available measures, estimates of intelligibility, stimulability testing, hearing screenings, and informal assessment procedures to diagnose disorders of articulation and phonology. The use of a quantitative severity measure is frequently omitted.

Among published single-word measures, the Goldman-Fristoe Test of Articulation (GFTA) (Goldman & Fristoe, 2000) and the Khan-Lewis Phonological Analysis (KLPA) (Khan & Lewis, 2002) were used most frequently among practicing clinicians. Skahan and colleagues (2007) speculate that the GFTA’s validity, reliability, versatility, and efficiency contribute to its continued use, although the information obtained from these measures may be too limited to determine appropriate goals for intervention. The SLPs who participated in the survey indicated that they administered and analyzed single-word tests more frequently than they elicited continuous speech samples. While single-word measures of articulation provide information about the child’s production of phonemes in all word positions, they provide neither typical nor optimal measures of speech production. For example, in a study of stimulus characteristics of single-word tests by Macrae (2017), he found that of the 12 popular tests of speech sound production studied, most included only on opportunity to demonstrate production of the majority of consonants. While McLeod and Baker (2014) included conversational speech sampling for the assessment of SSDs in their survey of Australian clinicians, they did not inquire about the use of a quantitative measure for assessing severity of involvement.

In addition, the psychometric properties of current single-word tests of speech sound production must also be considered. While progress has been made, most assessments still fail to demonstrate the full range of psychometric properties set forth by McCauley and Swisher in their 1984 study (Flipsen & Ogiela, 2015). With these weaknesses in mind, analysis of a continuous speech sample is recommended for quantitative measures of severity and a more complete representation of speech production (Morrison & Shriberg, 1992; Shriberg & Kwiatkowski, 1982a; Stoel-Gammon, 1987).

The preceding surveys examining common speech sound assessment practices among American and Australian SLPs indicated that a measure of severity of involvement is not commonly used among practicing clinicians, despite the existence of such measures. Although participants acknowledged the importance of intelligibility ratings, research has suggested that many clinicians rely on impressionistic estimates of intelligibility rather than conducting a more objective measurement (Skahan et al., 2007). Often, an impressionistic judgment as to the degree of impairment is made based on a continuum of severity ranging from mild to severe (Flipsen et al., 2005). The researchers found that clinicians tend to use judgments of both individual segments and whole words in making severity determinations and that perceptual severity ratings varied among even the most experienced clinicians. Clinical judgment appears to be the preferred practice pattern despite the availability of quantitative assessment measures of speech sound production that closely align with listeners’ perceptions of severity (Shriberg & Kwiatkowski, 1982b). As determination of severity of involvement is an area in which there is little consensus among professionals (ASHA, n.d.; Flipsen et al., 2005), the specific assessment procedures currently used by SLPs in the United States, particularly their knowledge and competence of the Percentage of Consonants Correct (PCC) metric, is of particular interest.

Percentage of Consonants Correct

The PCC metric is a quantitative severity measure developed by Shriberg and Kwiatkowski (1982a) designed to assess the severity of involvement in individuals with SSDs. The metric considers only the production of consonants, which are judged as correct or incorrect based on the PCC sampling rules. When calculating the PCC metric, the clinician determines the percentage of correctly articulated consonant sounds present out of all consonant sounds attempted during a 3- to 4-minutes conversational speech sample (Shriberg & Kwiatkowski, 1982a). According to Shriberg and Kwiatkowski (1982b), this percentage correlates to a corresponding severity level indicating the severity of the SSD. Individuals with a PCC greater than 85% receive a rating of mild, those with a score from 65%–85% receive a rating of mild-moderate, scores of 50%–64% receive a rating of moderate-severe, and individuals with PCC scores below 50% are classified as severe.

In addition, an extensive set of norms developed by Austin and Shriberg (1997) allow the individual’s performance to be compared to a normative sample of the same age, gender, and speech classification. For example, based on Austin and Shriberg’s norms, a 4-year-old boy with a mean PCC of 80.2% for all consonant sounds would be classified as exhibiting Normal Speech Acquisition (NSA); however, a 4-year-old boy with a mean PCC of 65% for all consonant sounds would be classified as Speech Delayed. As a quantitative measure, PCC may be valuable within the school setting as an interpretive tool to describe the severity of a child’s SSD to parents, teachers, and other professionals. In addition, quantitative severity measures derived from conversational speech samples may aid in determining eligibility for services in the public school system when single-word measures may not. As PCC is derived from a language sample, it would likely provide a more accurate representation of the severity of a child’s deficits within a more natural context.

Extension to the PCC Metric

To combat concern that the PCC may obscure important differences associated with certain types or subgroups of sounds Shriberg, Austin, Lewis, McSweeny, and Wilson (1997) developed the Percentage of Consonants Correct-Revised (PCC-R; Shriberg et al., 1997). When calculating the PCC-R, both common and uncommon clinical distortions (i.e., weak consonants, imprecise consonants and vowels, failure to maintain oral/nasal contrasts, and notable failure to maintain appropriate voicing) are scored as correct. While both are valuable metrics, the focus of the current investigation is on PCC and provides information on all three error types (i.e., substitutions, omissions, and distortions).

Clinician agreement in severity assessment

A comparison of phonological severity measures was conducted by Garret and Moran (1992) to determine the extent of agreement among PCC, phonological deviancy score (PDS), and the perceptual judgments of listeners. A total of 20 speech samples were obtained from children between the ages of 5 and 9, who were receiving treatment for phonological impairments ranging in severity from mild to severe. High correlations were observed between PCC and listener severity ratings, whereas a lower correlation was observed between PCC and PDS measures. The strong correlation between PCC, PDS, and listener severity ratings indicates that PCC and PDS are valuable and accurate indicators of severity (Garret & Moran, 1992).

Academic and Clinical Preparation

It is evident that quantitative measures are recommended for the diagnosis of SSD. However, the degree to which speech-language pathology students are being trained to use such measures, like PCC, is not clear. While research is growing with regard to the clinical and academic preparation needs of SLPs, a bulk of this literature investigates training in specialized areas, such as autism spectrum disorder (Plumb & Plexico, 2013), stuttering (Brisk, Healey, & Hux, 1997), and bilingualism (Hammer, Detweiler, Detweiler, Blood, & Qualls, 2004; Roseberry-McKibbin, Brice, & O’Hanlon, 2005). There is currently a lack of evidence, however, specific to an SLP’s training in the assessment of SSDs. Of importance is how much academic and clinical experience do SLPs receive with regard to ASHA’s preferred practice patterns in working with individuals with SSDs. As a measurement of severity is one such recommendation, it is of key interest to gain an understanding of both SLPs’ undergraduate and graduate training with regard to severity metrics, such as PCC, as well as their ability to correctly calculate this measure.

Purpose

The aim of this study is to investigate SLPs’ knowledge and competence regarding the use of PCC when assessing severity of involvement of a child’s SSD. Questions addressed SLPs’ education and training in the evaluation of child and adolescent SSDs and their familiarity with current research in this area, providing insight into whether adequate knowledge about PCC exists. Specifically, this study hopes to answer the following questions:

How much academic and clinical training in assessment and diagnosis of SSDs have SLPs received? It is hypothesized that survey participants will indicate extensive training in the assessment and diagnosis of SSDs.

What level of knowledge and competence do SLPs demonstrate in their determination of PCC? It is hypothesized that participants will lack accuracy when calculating PCC and they will lack a comprehensive understanding of the PCC rules.

Method

Participants

A total of 62 respondents completed the survey. To meet inclusion criteria, respondents held ASHA certification or were clinical fellows with experience working with children. If inclusion criteria were not met, they were taken to the end of the survey, and their responses were not used for analysis. The majority of respondents indicated that their highest degree received was a master’s degree (90.3%; n = 56), while 9.6% (n = 6) earned a doctorate degree. With regard to the location in which the respondents are practicing and professionally licensed, 61 (98.4%) indicated that they are licensed in the United States, while one individual (1.6%) reported holding licensure in Canada. Participants were practicing in 22 states and represented the demographic regions of the Northeast, South, Midwest, and West. Of the 59 respondents who indicated the location of their licensure, the largest percentage of respondents practiced in the South (n = 21; 35.6%), followed by the West (n = 17; 28.8%), Northeast (n = 11; 18.6%), and Midwest (n = 10; 16.9%). Data related to professional experience and caseload characteristics are found in Table 1.

Table 1.

Professional Experience and Caseload Characteristics.

Background	%	n
Highest degree received
Master’s degree	90.3	56
Doctorate degree	9.7	6
Date highest degree received
Prior to 1985	11.3	7
1985–1994	24.2	15
1995–2004	16.1	10
2005 to present	48.4	30
Years of professional experience, beginning with CFY
<1 year	3.2	2
1–5 years	17.7	11
6–10 years	25.8	16
11–15 years	4.8	3
16+ years	48.4	30
Percentage of caseload comprised of children
0–10	1.6	1
11–20	0	0
21–30	0	0
31–40	0	0
41–50	0	0
51–60	0	0
61–70	1.6	1
71–80	0	0
81–90	4.8	3
91–100	91.9	57
Percentage of caseload receiving treatment for speech sound disorders
0–10	16.1	10
11–20	8.1	5
21–30	11.3	7
31–40	11.3	7
41–50	12.9	8
51–60	9.7	6
61–70	11.3	7
71–80	11.3	7
81–90	1.6	1
91–100	6.5	4

Note. CFY = clinical fellowship year; N = 62; n = number of respondents; % = percentage of respondents.

Materials

To answer the questions of the study, the investigators created a web-based, 33-item survey via Qualtrics software to address questions in five main areas: (a) general demographic information; (b) educational background; (c) service delivery; (d) knowledge regarding PCC; and (e) competence and confidence in calculating PCC. Following is a description of each area.

Part I questions were designed to acquire information about SLPs’ general professional background, years of experience, and the age range of their current caseload. Part II questions addressed the participants’ educational background with regard to the nature and assessment of SSDs. Pre-professional clinical experience in the assessment of children with SSDs was also explored in addition to questions querying professional continuing education (CE) courses related to SSDs. Part III questions addressed participants’ familiarity with the rules used to calculate PCC. Participants were asked to answer a set of true/false questions based on the rules for calculating PCC defined by Shriberg and Kwiatkowski (1982b). Part IV questions addressed participants’ ability to accurately calculate PCC. Participants were instructed to calculate PCC for four speech samples of increasing complexity; they were then asked to determine the presence of a speech delay and assign an appropriate severity rating. In addition to their competence with the measure, participants were asked to indicate their confidence in their ability to calculate PCC.

Procedure

Prior to the distribution of the survey, an initial draft was piloted by articulation and phonology faculty members from six universities who provided feedback about the survey’s content, and format changes were made accordingly. Participants in this study were recruited via three methods. First, the following ASHA Special Interest Groups were selected, due to their focus on service provision to school-aged individuals and the resulting probability that many of their members may be working with individuals with SSDs: Language Learning and Education Special Interest Group (SIG 1), Craniofacial and Velopharyngeal Disorders Special Interest Group (SIG 5), and School-Based Issues Special Interest Group (SIG 16). An introductory email, containing an information letter with a weblink to the online survey, was sent to the coordinators of these divisions and then posted on the respective listservs.

Second, a search of the ASHA membership directory through the ASHA community was conducted to find potential participants who (a) hold ASHA certification in SLP; (b) listed their current place of work as a school or other location where the predominant population is pediatric; and (c) were employed in one of nine states: Alabama, Colorado, Florida, Illinois, Louisiana, Massachusetts, Missouri, New York, and Washington. This selection process was designed to ensure equal geographic representation by selecting one state from each geographic division of the United States as defined by the U.S. Census (U.S. Census Bureau, 2010). The researchers emailed the same information letter that was posted on the listservs to every fifth individual who met study criteria as listed in the membership directory. A total of 100 emails were sent to SLPs in each state. Finally, a message containing a brief description of the survey, its purpose, and an embedded link to direct interested individuals to the survey was posted on the ASHA Facebook page.

The survey was administered using the online survey tool Qualtrics, which is a secure, Internet-based software program. All data were collected anonymously. A total of 900 emails were sent via the ASHA membership directory, and a total of 130 individuals initiated the survey. Responses were then filtered for survey completion, resulting in a total of 62 responses. A total of 10 participants (16.1%) were recruited via the ASHA membership directory, yielding a response rate of 1.1% for that recruitment method. While response rates could not be calculated for the remaining recruitment methods, 10 (16.1%) individuals were recruited via the ASHA Community Discussion Board; 1 (1.6%) individual was recruited via the ASHA Facebook page; 5 (8.1%) participants were contacted via Special Interest Group 5, Craniofacial and Velopharyngeal Disorders; 31 (50.0%) participants were recruited via ASHA Special Interest Group 16, School-Based Issues; and 5 (8.1%) indicated “other.”

Results

Educational Background With Regard to Assessment of SSDs

When asked how many courses in their graduate studies addressed speech sound assessment as a component of the course, most respondents (96.8%; n = 60) reported at least one graduate course that met criteria. Although 3.2% of respondents (n = 2) indicated that they had no graduate coursework addressing the assessment of SSDs, it is possible that these individuals had an undergraduate course that addressed speech sound assessment as a component of the course. Of these courses, the number of contact hours of instruction devoted to speech sound assessment ranged from 1–3 hours (22.4%; n = 13) to 9 or more hours (27.6%; n = 16), with 15 respondents (25.8%) reporting 4–6 hours and 14 (24.1%) reporting 7–9 hours of instruction. When asked about CE courses, 83.9% of respondents (n = 52) indicated that they had obtained CE credits related to SSDs since receiving their highest degree. In addition, 57.7% (n = 30) of 52 respondents reported that they annually seek out CE opportunities related to SSDs.

With regard to the level of exposure participants had to PCC, the majority of the 58 respondents who answered this question reported they had at least some exposure during graduate study (56.9%; n = 33). Slightly more than one third of the participants responded that they had experience calculating PCC either through practice in a course or with a client (36.2%; n = 21); however, 18.9% reported they did not have exposure to PCC in graduate study, and 24.1% of respondents did not recall their exposure to the PCC metric.

Respondents who indicated they had exposure to PCC in undergraduate or graduate school were then queried as to the amount of time devoted to the measure. Of the 33 respondents who answered this question, responses ranged from no classes to more than 1 month. A total of 31 respondents (94.0%) indicated that at least one class was devoted to PCC, while 6.1% (n = 2) reported that PCC was not discussed at all. Respondents were also presented with a Likert-type scale designed to determine whether they felt they could have benefited from additional coursework devoted to speech sound assessment. In response to this question, 14 participants (22.6%) indicated that they strongly agreed, 18 (29.0%) agreed with the statement, 17 (27.4%) neither agreed nor disagreed, 11 participants (17.7%) disagreed, and 2 (3.2%) strongly disagreed with the statement.

Respondents were then asked a series of questions to determine the level of clinical experience received during their graduate training programs with regard to the assessment of SSDs. When asked how many children with SSDs they had the opportunity to assess during their practicum in the university clinic, 91.8% (n = 56) of the 61 respondents (n = 56) indicated that they had the opportunity to assess at least one child with an SSD. When asked if they had the opportunity to use PCC in the university clinic to assess the severity of a child’s SSD, most (80.3%; n = 49) of the 61 respondents indicated that they did not. Respondents were also queried as to the number of children with SSDs they had the opportunity to assess during their off-campus practicum. Of the 61 respondents who had an off-campus practicum experience, 91.8% (n = 56) indicated that they had assessed at least one child with an SSD. Of those respondents who did have an off-campus practicum experience, only 18.0% (n = 11) reported having the opportunity to use PCC to assess the severity of a child’s SSD.

Participants were then presented a Likert-type scale designed to determine whether they believed they could have benefited from additional clinical experience related to assessing children with SSDs. When asked to describe the degree to which they agreed with the previous statement, 16.1% of individuals (n = 10) strongly agreed, 40.3% (n = 25) agreed, 24.2% (n = 15) neither agreed nor disagreed, 16.1% (n = 10) disagreed, and 3.2% (n = 2) strongly disagreed. Most participants (56.5%; n = 35), therefore, either agreed or strongly agreed that they would have benefited from additional clinical experience assessing SSDs in children.

Knowledge of PCC Rules

Participants were asked a series of true/false questions designed to determine their knowledge of the rules used to calculate PCC. It should be noted that two queried rules were listed twice and therefore were omitted from data analysis. Overall, participants demonstrated strengths and weaknesses in their knowledge of the rules of PCC Calculation (see Table 2). The largest percentage of respondents demonstrated understanding that deletions, substitutions, and distortions would be scored as incorrect. It is possible that survey respondents correctly identified deletions, substitutions, and distortions as incorrect due to their prevalence in children with SSDs (Preston & Edwards, 2010; Smit, 1993). However, the minority of respondents demonstrated understanding of rules, such as adding a consonant before a vowel, deletion of stressed and unstressed /ɝ/ and /ɚ/, and deletion of initial /h/ and final /n/ for /ŋ/ substitution; these rules are more complex, and the errors themselves perhaps less common than deletions, simple substitutions of target consonants, and consonant distortions.

Table 2.

SLPs’ Knowledge of the Rules Governing PCC.

	Correct	Incorrect
True/False Questions	n (%)	n (%)
*Consonants in the third or successive repetitions of adjacent words should be scored only if articulation changes (e.g., /kʌp, kʌp, kʌp/ versus /kʌp, tʌp, kʌp/)	25 (60.9)	16 (39.0)
*Deletion of a target consonant is scored as incorrect	37 (90.2)	4 (9.8)
*Substitution of another sound for a target consonant is scored as incorrect	38 (92.7)	3 (7.3)
Addition of a consonant before a vowel (e.g., /hæt/ for “at”) is scored as incorrect	13 (31.7)	28 (68.3)
*Partial voicing of initial target consonants is scored as incorrect	26 (63.4)	15 (36.6)
*Distortions of a target consonant are scored as incorrect	36 (87.8)	5 (12.2)
*Addition of a sound to a correct or incorrect target consonant (e.g., /kɑrks/ for “cars”) is scored as incorrect	29 (70.7)	12 (29.3)
Deletion of stressed and unstressed /ɝ/ and /ɚ/ is scored as incorrect	14 (34.1)	27 (65.9)
*Deletion of initial /h/ and final /n/ for /ŋ/ substitution is scored as incorrect only when they occur in stressed syllables	15 (36.6)	26 (63.4)
*Deletion of post-vocalic /r/ is scored as incorrect	28 (68.3)	13 (31.7)
Allophones are scored as incorrect (e.g., /bʌɾɚ/ for /bʌtɚ/)	23 (56.1)	18 (43.9)

Note. True answers are marked with an asterisk; n = number of participants; % = percentage of participants. SLPs = speech-language pathologists; PCC = Percentage of Consonants Correct.

Calculation of PCC

To determine their competence in using the PCC metric, participants were asked to calculate PCC for four speech samples of increasing complexity. Both the child’s productions and the adult glosses were provided for calculation. Of the 62 individuals who completed the survey, 41 participants elected to respond to questions in this section. When asked to calculate the PCC for a single word, 34.1% of participants (n = 14) provided an accurate percentage; however, 60.9% (n = 25) calculated a PCC that was within the appropriate severity range as proposed in Shriberg and Kwiatkowski (1982a). When asked to calculate the PCC for a phrase, 31.7% of participants (n = 13) provided an accurate percentage, while 41.5% (n = 17) calculated a PCC that was within the appropriate severity range. When asked to calculate the PCC at the sentence level, 12.2% of participants (n = 5) provided an accurate percentage, while 63.4% (n = 26) calculated a PCC that was within the appropriate severity range. When asked to calculate the PCC for a three-sentence speech sample produced by a 4-year-old female speaker of Standard American English (SAE), 9.8% of participants (n = 4) provided an accurate percentage, while 34.1% (n = 14) calculated a PCC that was within the appropriate severity range. Participants’ calculations of PCC were more accurate when provided shorter utterances with simpler errors, such as consonant substitutions. However, participants had more difficulty with utterances of increasing length and demonstrated decreased knowledge of more complex rules, including those regarding the scoring of syllable repetitions and the addition of a consonant before a target vowel.

Participants were then asked to determine the presence of a speech delay based on the PCC for a given sample whose PCC of 66.6% correlated to a mild-moderate severity rating as proposed by Shriberg and Kwiatkowski in 1982. After calculating the PCC for a three-sentence sample produced by a 4-year-old female speaker of SAE, 65.0% of 40 respondents (n = 26) correctly identified the presence of a speech delay. The 26 individuals who correctly identified the sample as being indicative of a speech delay were then queried as to level of severity. A total of 25 respondents (96.2%) appropriately placed the child’s SSD within Shriberg and Kwiatkowski’s mild-moderate range, with both mild and moderate responses considered correct. A total of 50.0% of respondents (n = 13) indicated a mild disorder, 46.2% (n = 12) indicated a moderate disorder, and 3.8% (n = 1) indicated a severe disorder (Shriberg & Kwiatkowski, 1982a). When asked what they based their severity rating on, 26.9% of respondents (n = 7) indicated that they based their judgment of severity on their PCC calculation, while 73.1% (n = 19) based their judgment of severity on clinical experience. Finally, participants were asked to rate their degree of confidence using a 5-point Likert-type scale. When asked to indicate the extent to which they agreed with the statement, “I am confident in my ability to calculate PCC,” 29.3% (n = 12) of respondents strongly disagreed, 24.4% (n = 10) disagreed, 21.9% (n = 9) neither agreed nor disagreed, 21.9% (n = 9) agreed, and 2.4% (n =1) strongly agreed.

Discussion

The purpose of this study was to explore SLPs’ knowledge and competence regarding PCC, specifically those who serve school-aged populations. The findings of the study indicate that most SLPs have limited knowledge with regard to the rules used to calculate PCC, and their resulting ability to accurately calculate PCC is somewhat low. As a quantitative measure of severity is advocated by ASHA (n.d.), these findings suggest that SLPs may benefit from increased clinical and academic training supporting the initial hypotheses; however, the possibility exists that the development of a more efficient quantitative tool for determining speech sound severity may be required to sufficiently address the need indicated by the current findings.

Educational Background With Regard to Assessment of SSDs

Nearly all respondents reported that they received graduate-level training in the assessment of SSDs, during which they were exposed to both nonstandardized and standardized assessment procedures. As the GFTA and the KLPA have been found to be the most commonly used standardized assessment tools (Skahan et al., 2007), respondents’ high familiarity with these measures may be reflective of this. The possibility exists that the instruments which clinicians become the most familiar with in graduate training will become used the most clinically. Equally possible, Skahan et al. (2007) speculated about its clinical popularity that the GFTA’s and KLPA’s validity, reliability, versatility, and efficiency contribute to their high frequency of exposure in the graduate curriculum. In addition, knowing that clinicians are most likely to use this test throughout their careers, it is logical that the large majority of programs would give students exposure to this often-used assessment.

Exposure to PCC

While slightly over half of survey respondents discussed PCC as a component of their graduate coursework, only slightly more than one third of participants had experience calculating PCC, either through practice in a course or with a client. The majority of participants also reported limited experience related to using PCC as a speech sound assessment practice in their clinical practicum, with only around 20% having experience using PCC to assess speech sound severity. It was unexpected that participants did not report more exposure to PCC in their undergraduate and graduate training programs, as objective measures of severity of involvement like PCC are specifically mentioned in ASHA’s online Practice Portal for articulation and phonology. It is also interesting that, although their exposure to PCC was limited, participants recognized the importance of the measure as slightly over half of the participants believed that they could have benefited from additional academic and clinical training in the assessment of SSDs in children.

Knowledge of PCC Rules

Calculation of PCC requires knowledge of a complex set of sampling and scoring rules. Participants in this study demonstrated inconsistent knowledge with regard to the rules used to calculate PCC. While typically more than half of respondents demonstrated an understanding of the rules regarding how to distinguish correct from incorrect utterances, respondents were most familiar with the commonly discussed and encountered errors of substitutions, distortions, and omissions. They demonstrated greater difficulty with more complex and perhaps less encountered errors such as consonant epenthesis; when presented, participants demonstrated a tendency to incorrectly label them as errors. The inconsistent knowledge of the rules for calculating PCC was not surprising, given that most participants indicated limited exposure to PCC during their undergraduate and graduate training programs, as well as limited opportunities to use PCC to assess speech sound severity during their clinical practicum experiences. Lack of exposure and an incomplete knowledge of the extensive set of rules governing the use of PCC may contribute to clinicians’ limited use of the measure.

Competence and Confidence in Calculation of PCC

Of the 62 individuals who completed the survey, only 41 participants elected to respond to questions requiring the application of PCC to samples of varying length. The authors believe that the lack of academic/clinical exposure and general lack of practical use of this measure may have negatively affected the response rate for questions in this section. In keeping with the initial hypotheses of the current investigation, most respondents calculated PCC incorrectly. Results of the study also suggested that as the length of the speech sample increased, participants’ ability to calculate an accurate PCC decreased. When asked to calculate the PCC for a single word, slightly over one third of respondents provided an accurate percentage, as opposed to the one tenth of respondents who provided an accurate PCC for a three-sentence speech sample. As spontaneous speech samples are typically well beyond one-word utterances, this is of concern; however, results were more promising when analyzed based on the participants’ ability to calculate a PCC within the appropriate severity range as proposed by Shriberg and Kwiatkowski (1982b). These findings indicated that a larger percentage of respondents were able to identify the correct severity level of a child based on PCC calculation in instances where the calculation was not exact. Even through this lens, however, at least one third of clinicians were not able to do so correctly at the word, phrase, or sentence level. Factors involved in this may be lack of academic or clinical exposure, decreased knowledge of the rules, or simply that they do not frequently use the measure so have not retained the skills necessary to accurately calculate PCC.

After providing a judgment of speech sound severity, almost three quarters of participants indicated that their judgment of severity was based on clinical experience rather than determination of PCC. These findings were somewhat expected in light of participants’ lack of academic and clinical experience using PCC, as well as their frequent omission of the measure from their speech sound assessment service delivery. Not surprisingly, over half of the participants in this study indicated that they were not confident in their ability to calculate PCC. The lack of confidence felt by participants may have negatively impacted their likeliness to respond to the questions requiring them to calculate the PCC for a given sample.

Limitations and Future Directions

The primary limitation of the current investigation is the low response rate. Of the 900 individuals invited to participate in the survey via email, 10 completed the survey through this recruitment method, yielding a 1.1% response rate. The low response rate indicates that this sample may not fully represent the general population. Another factor that could limit the current investigation’s generalization is the possibility that only SLPs interested in speech sound assessment and PCC may have responded to the survey. If an SLP has a particular interest in this subject, they may regularly seek out additional information on the topic and be more aware than an SLP who does not have a particular interest in the subject. In fact, more than half of the participants in this study indicated that they seek out CE credits related to SSDs annually. Therefore, the possibility exists that clinicians are not as knowledgeable about the PCC measure as indicated by the results of this survey.

There is currently a lack of consensus among practicing clinicians regarding the best methods of determining speech sound severity in children (Flipsen et al., 2005). As PCC is a well-validated measure that closely aligns with listener perceptions of severity, it is surprising that the tool is rarely used in current speech sound assessment practices (McLeod & Baker, 2014; Skahan et al., 2007). It would be of interest to qualitatively examine the reasons behind clinicians’ limited use of the measure to assess speech sound severity such as time constraints, level of difficulty, lack of confidence, and lack of academic and clinical training.

Clinical Implications

The results of this study are consistent with recent findings (McLeod & Baker, 2014; Skahan et al., 2007) regarding current speech sound assessment practices. Participants in this study frequently relied on standardized assessments such as the GFTA and KLPA, as well as phonetic transcription, phonetic inventory, severity ratings, and intelligibility ratings. However, quantitative severity measures, particularly PCC, are rarely used despite being well-validated and closely aligned with listener perceptions of severity. In addition, participants reported limited academic and clinical exposure to PCC, inconsistent knowledge of PCC rules, and a lack of confidence using the measure; therefore, it appears that some pediatric SLPs would benefit from additional instruction with regard to PCC rules and clinical experience calculating PCC.

Although additional clinical and academic exposure to PCC may increase SLPs’ level of familiarity and competence with the measure, it does not necessarily follow that this increased knowledge will result in SLPs’ more frequent use of the measure in their speech sound assessment practices. Use of the PCC metric requires transcription skills and can be a time-consuming task. Although increased experience and practice with the measure may increase the speed with which an individual can complete this assessment, some clinicians may not consider the use of PCC to be a feasible task given the limited assessment time available to many clinicians, particularly those in the public school setting, potentially rendering it more appropriate for research (Khan, 2002). Despite its potential lack of practicality within the school setting, clinicians’ omission of PCC and other objective severity measures is concerning. In light of ASHA’s recommendations, the findings of this study suggest that a need exists among SLPs working with school-aged children. Yet to be determined is whether this need can be sufficiently addressed by increased exposure to PCC during academic and clinical training in the area of speech sound assessment or if there is a need for a more expedient process by which to objectively quantify severity.

It is possible that the time-consuming task of hand-calculating PCC could be made more efficient and clinically practical through the use of an automatized metric. For example, the computer program Phon (Hedlund & Rose, 2019) is a free software program that allows the user to phonetically transcribe language samples using an IPA map and then run a query to calculate PCC. This method requires the collection of a conversational speech sample and transcription of both the orthographic and phonetic gloss into the program. As the participants in this study indicated that they received training and exposure to phonetic transcription and often use phonetic transcription to assess the speech of children on their caseloads, this may be prove to be a solution; however, it should be noted that use of such a program necessitates a certain level of computer ability and will require time to learn how to use the program. Use of a program, such as Phon, would perhaps negate the effects of clinicians’ decreased academic and clinical exposure to the measure during their graduate and undergraduate training, including decreased knowledge of the complex sampling and scoring rules governing PCC in addition to decreased competence in calculation of PCC. It is possible that a program which increases the ease with which PCC is calculated could increase SLPs’ use of such a highly validated, objective measure of severity.

Footnotes

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

References

American Speech-Language-Hearing Association. (2004). Preferred practice patterns for the profession of speech-language pathology [Preferred practice patterns]. Retrieved from www.asha.org/policy

American Speech-Language-Hearing Association. (2018). ASHA 2018 schools survey: SLP caseload characteristics report. Retrieved from https://www.asha.org/uploadedFiles/2018-Schools-Survey-Caseload-Trends.pdf

American Speech-Language-Hearing Association. (n.d.). Speech sound disorders: Articulation and phonology (Practice Portal). Retrieved from www.asha.org/Practice-Portal/Clinical-Topics/Articulation-and-Phonology

Austin

Shriberg

L. D.

(1997). Lifespan reference data for ten measures of articulation competence using the speech disorders classification system (SDCS). Madison: Waisman Center, University of Wisconsin–Madison. Retrieved from http://www2.waisman.wisc.edu/phonology/techreports/TREP3_REV.PDF

Brisk

D. J.

Healey

E. C.

Hux

K. A.

(1997). Clinicians’ training and confidence associated with treating school-age children who stutter: A national survey. Language, Speech, and Hearing Services in the Schools, 28, 164–176.

Farquharson

Boldini

(2018). Variability in interpreting “educational performance” for children with speech sound disorders. Language, Speech, and Hearing Services in Schools, 49, 938–949.

Flipsen

Jr. Ogiela

D. A.

(2015). Psychometric characteristics of single-word tests of children’s speech sound production. Language, Speech, and Hearing Services in Schools, 46, 166–178.

Flipsen

Jr. Hammer

J. B.

Yost

K. M.

(2005). Measuring severity of involvement in speech delay: Segmental and whole-word measures. American Journal of Speech-Language Pathology, 14, 298–312.

Garret

K. K.

Moran

M. J.

(1992). A comparison of phonological severity measures. Language, Speech, and Hearing Services in Schools, 23, 48–51.

10.

Goldman

Fristoe

(2000). Goldman-Fristoe Test of Articulation. Circle Pines, MN: American Guidance Service.

11.

Hammer

C. S.

Detweiler

J. S.

Detweiler

Blood

G. B.

Qualls

C. D.

(2004). Speech-language pathologists’ training and confidence in serving Spanish–English bilingual children. Journal of Communication Disorders, 37, 91–108.

12.

Hedlund

Rose

(2019). Phon 3.0 [Computer Software]. Available from https://phon.ca

13.

Khan

L. M.

(2002). The sixth view: Assessing preschoolers’ articulation and phonology from the trenches. American Journal of Speech-Language Pathology, 11, 250–254.

14.

Khan

L. M.

Lewis

N. P.

(2002). Khan-Lewis Phonological Analysis. Circle Pines, MN: American Guidance Service.

15.

Macrae

(2017). Stimulus characteristics of single-word tests of children’s speech sound production. Language, Speech, and Hearing Services in Schools, 48, 219–233.

16.

McCauley

R. J.

Swisher

(1984). Psychometric review of language and articulation tests for preschool children. Journal of Speech and Hearing Disorders, 49, 34–42.

17.

McLeod

Baker

(2014). Speech-language pathologists’ practices regarding assessment, analysis, target selection, intervention, and service delivery for children with speech sound disorders. Clinical Linguistics & Phonetics, 27, 508–531.

18.

Morrison

J. A.

Shriberg

L. D.

(1992). Articulation testing versus conversational speech sampling. Journal of Speech and Hearing Research, 35, 259–273.

19.

Plumb

Plexico

(2013). Autism spectrum disorders: Experience, training, and confidence levels of school-based speech-language pathologists. Language, Speech, and Hearing Services in Schools, 44, 89–104.

20.

Preston

Edwards

M. L.

(2010). Phonological awareness and types of sound errors in preschoolers with speech sound disorders. Journal of Speech, Language, and Hearing Research, 53, 44–60.

21.

Roseberry-McKibbin

Brice

O’Hanlon

(2005). Serving English language learners in school settings: A national survey. Language, Speech, and Hearing Services in Schools, 36, 48–61.

22.

Shriberg

L. D.

Austin

Lewis

B. A.

McSweeny

J. L.

Wilson

D. L.

(1997). The percentage of consonants correct (PCC) metric: Extensions and reliability data. Journal of Speech, Language, and Hearing Research, 40(4), 708–722.

23.

Shriberg

L. D.

Kwiatkowski

(1982a). Phonological disorders I: A diagnostic classification system. Journal of Speech and Hearing Disorders, 47, 226–241.

24.

Shriberg

L. D.

Kwiatkowski

(1982b). Phonological disorders III: A procedure for assessing severity of involvement. Journal of Speech and Hearing Disorders, 47, 256–270.

25.

Skahan

S. M.

Watson

Lof

G. L.

(2007). Speech-language pathologists’ assessment practices for children with suspected speech sound disorders: Results of a national survey. American Journal of Speech-Language Pathology, 16, 246–259.

26.

Smit

A. B.

(1993). Phonologic error distributions in the Iowa-Nebraska articulation norms project: Consonant singletons. Journal of Speech and Hearing Research, 36, 533–547.

27.

Stoel-Gammon

(1987). Phonological skills of 2-year-olds. Language, Speech and Hearing Services in Schools, 18, 323–329.

28.

Tuomi

Ivanoff

(1977). Incidence of speech and hearing disorders among kindergarten and Grade 1 children. Special Education in Canada, 51, 5–8.

29.

U.S. Census Bureau. (2010). Census 2010. Washington, DC: U.S. Department of Commerce.