Parental Judgments of Preschoolers' Speech and Language Development: A Resource for Assessment and IEP Planning

An important component of a preschooler's speech and language evaluation is the parental report of the child's developmental history and current levels of functioning. Often parents are asked to share any particular concerns they have about their child's communicative development. Some parents remark on their child's speech, “My child has blurry speech,” or “I can understand what my child says, but no one else can.” Others focus on their child's limited expressive language skills: “My child seems to understand what I say to him, but he doesn't build many of his own sentences.”

Such replies provide the speech-language pathologist (SLP) with a rough idea of how a parent perceives the child's communicative limitations, but it is usually difficult to know what to make of such comments. Because parents and SLPs usually do not use the same terminology, there is no assurance that they are referring to the same communicative phenomena, and there is little information available about how accurate parents are likely to be with such general characterizations of preschoolers' communicative competencies.

On the other hand, valid parental reports would be of great value to the SLP. Such information could help the SLP evaluate the validity of the formal test scores obtained in those domains for which tests are available (i.e., speech, semantics, and syntax). Furthermore, valid parental reports could provide valuable information about conversational skills not readily observable in formal instructional settings or communicative abilities evident in situations familiar to the child. Finally, if parents and SLPs could share their observations in a way that was meaningful to each of them, it would greatly facilitate intervention planning, formulation of goals for the child's Individualized Educational Plan (IEP), and evaluation of change over the period of intervention.

In this article we report on an instrument, the Speech and Language Assessment Scale (SLAS), which was designed to elicit parental judgments of preschool children's speech and language skills in a way that can be useful for initial assessment, for development of IEP goals and intervention plans, for monitoring a child's speech and language abilities, and for ongoing consultation with parents about their child's communicative capabilities. We will proceed in the following way. After a brief review of methods of parent report, we describe the development of SLAS, with information about the instrument's reliability and validity. We then report the results of a study of parent and SLP ratings of children, identifying the areas of agreement and disagreement. We conclude with clinical implications and recommended uses of SLAS.

METHODS OF PARENT REPORT

There are three general methods for obtaining information from parents (Lichtenstein & Ireton, 1984). The method most frequently used is a comprehensive questionnaire. This is generally completed prior to, or as part of, the initial diagnostic assessment. Information obtained from comprehensive questionnaires includes the child's birth and medical history, the age at which the child attained various developmental milestones (e.g., walked, spoke first words), and the family's history of speech or language disorders. The second approach is to conduct an interview with the parent. Interviews may be guided by standardized instruments such as the Verbal Language Development Scale (Meacham, 1971) or the Vineland Adaptive Behavior Scales (Sparrow, Balla, & Cicchetti, 1984). One limitation of standardized interviews is that only a few behaviors are sampled at each developmental level. Alternatively, professionals may conduct parent interviews in a relatively open-ended manner. This allows parents to describe their child's communication skills in great detail, but makes it more difficult for the professional to repackage this information and identify the parents' greatest areas of concern. The third method for obtaining information from parents is to have them describe their child's current level of functioning, guided by a standardized developmental inventory. Parents' descriptions of current functioning are believed to be more reliable than retrospective reports of what the child was doing at 12 or 18 months of age (Dale, Bates, Resnick, & Morisset, 1989). In addition, inventories can be collected at periodic intervals as an indication of the parent's perception of a child's progress.

Recently, two instruments utilizing parent report have been developed to assess children's early language development. Rescorla (1989) developed and evaluated the Language Development Survey (LDS), a vocabulary checklist which contains 309 words arranged in 14 categories (e.g., food, animals, actions). Parental reports of their 2-year-old children's total vocabulary score on the LDS were highly correlated with standardized measures of language development (see also Rescorla, this issue). The validity of parental reports was also documented by Dale (1991) for the MacArthur Communicative Developmental Inventory: Toddler (CDI). The CDI assesses children's vocabulary and syntactic development. The first section contains 680 words arranged in 22 categories. The second section assesses syntactic development. For example, parents are asked to identify whether their child uses word endings (e.g., plural s, progressive ing). To assess syntactic complexity, parents are asked to mark the utterance “that sounds MOST like the way your child talks right now.” Dale established strong evidence of validity between these CDI measures obtained from parent report and standardized measures of vocabulary and expressive language development in a sample of 24 children at 24 months of age.

These two instruments demonstrate that parent report can be a valid method of obtaining information for toddlers' vocabulary and syntactic development. In fact, parent report is an integral component of most assessments of very young children, that is, those under 30 months of age, and such reports are widely recognized as essential for accurate assessments. As children approach preschool, however, systematic efforts to elicit information via parent report are much less common. It was our intent to develop an instrument that would allow parents to participate in ongoing assessments of their preschooler's communication skills.

For this reason, SLAS was developed (Rice, Wilcox, Liebhaber, & Hadley, 1989). It was field-tested in the Language Acquisition Preschool (LAP) at the University of Kansas. LAP is an integrated preschool designed to facilitate language development for children with various degrees of language proficiency. Three groups of children are served in this setting: children developing language normally, children with language impairments, and children learning English as a second language. At any one time, about 36 children are enrolled, equally distributed across the three groups. LAP is affiliated with a local school district and meets all state and federal service regulations for the development of IEPs and other intervention procedures. Two adults are in the classroom on a regular basis, the Lead Teacher and an Assistant Teacher. Field testing was carried out over a period of 3 years.

#

THE SPEECH AND LANGUAGE ASSESSMENT SCALE

The Speech and Language Assessment Scale is a brief questionnaire made up of 19 individual items. The items address a range of communicative dimensions. Some items correspond to formal measures of speech or language, such as articulation, receptive or expressive vocabulary, or mean length of utterance (MLU). Items that describe a child's conversational assertiveness or responsiveness were also developed. Finally, communication skills that preschool and kindergarten teachers deemed necessary for school readiness were also included on the questionnaire (Hains, Fowler, Schwartz, Kottwitz, & Rosenketter, 1989). Overall, the SLAS asks for parental judgments of a child's articulation and general intelligibility, vocabulary, sentence construction, and conversational skills using a seven-point Leikart scale (see Appendix).

Reliability

Interrater reliability for the instrument was assessed for individual SLAS items over a period of five semesters. Ratings were obtained from the LAP Lead Teacher, who was also a certified SLP, and the LAP Assistant Teacher. Both individuals completed the questionnaires independently for each child enrolled in LAP. Interrater reliability was assessed using the ratings of native speakers of English only. The number of children rated each semester varied from 13 to 20.

Individual analyses of variance (ANOVA) were conducted for each of the items, to test for possible differences between raters. No significant differences were found for any of the 19 SLAS items; however, only 14 of the 19 items were significantly correlated between raters across four or more semesters. The remaining five items proved to be unreliable (See Hadley, Rice, & Wilcox, in preparation; Hadley, Rice, & Wilcox, 1990; Rice & Wilcox, 1990). Three of these items reflected comprehension abilities (i.c., understanding what others say, understanding directions, following directions). The other two items addressed children's cultural awareness and whether or not they spoke too loud or soft. Because these items proved unreliable between teachers observing the children in the same setting, they were dropped from the remaining analyses.

#

Validity

The 14 items that met reliability criteria were then placed in six scales reflecting the following areas: assertiveness, responsiveness, semantics, syntax, articulation, talkativeness. This discussion is limited to the first five scales. The individual items included in these scales are listed in [Table 1].

Table 1
Individual Items Contained in Composite Scales
Scale		Items
Assertiveness	1.	asks questions properly
	11.	gets what (s)he wants by talking
	12.	starts conversation with peers
Responsiveness	2.	answers questions properly
	13.	keeps conversations going with peers
Semantics	5.	number of words known
	6.	uses words properly
	10.	uses proper words when talking
Syntax	14.	length of sentences
	15.	makes “grown-up” sentences
Articulation	4.	says sentences clearly
	7.	gets message across when talking
	16.	says sounds in individual words correctly

Discriminant function analysis procedures were used to evaluate the effectiveness of these five scales in correctly classifying children into the appropriate language groups: normally developing language, language- impaired, speech- (and language-) impaired. Stepwise variable selection was used to examine the composite scales that entered into the discriminant functions for each of the five semesters. Three scales emerged as most important for predicting group membership: articulation, assertiveness, and semantics. Using these scales alone, the discriminant functions correctly classified an average of 86% of the children in each sample (range = 75% to 95%).

Finally, Weinberg (1991) examined the construct validity of the SLAS. The mothers of 42 preschoolers rated their children's speech and language competencies, along with a certified SLP. The mean scores for the composite SLAS scales were compared with standardized scores on a battery of formal speech and language measures: (a) Reynell Developmental Language Scales—Revised (Reynell, 1985); (b) Peabody Picture Vocabulary Test—Revised (Dunn & Dunn, 1981); (c) Goldman-Fristoe Test of Articulation (Goldman & Fristoe, 1986); (d) mean length of utterance (MLU) following the conventions of Miller (1981); and (e) descriptive measures of verbal interaction obtained from the Social Interactive Coding System (Rice, Sell, & Hadley, 1990). The composite SLAS scales were found to have moderate to moderately high correlations for all comparisons of interest for both mothers and the SLP. Interestingly, the highest correlation between maternal ratings and formal measures was for the Articulation composite and the Goldman-Fristoe Test of Articulation, whereas the highest correlation between the SLP's ratings and formal measures was that between the Syntax composite and MLU.

#

Intended Use of the SLAS

These initial analyses suggest that the SLAS is a reliable and valid tool. It is also able to discriminate between children with age-appropriate language skills and those with speech and/or language disorders. However, it is intended to supplement, not replace, formal diagnostic evaluations. It is designed to be used by parents to provide ongoing assessments of their child's speech and language competencies. With this information, an SLP will be better prepared to help parents participate more fully in the process of formulating their child's Individualized Education Program (IEP). For example, when parents and the SLP are in close agreement, the SLAS could be used to select appropriate goals for intervention. If there are differences in perspective between the parents and the SLP, individual items can serve as a point of departure for discussing the factors or social contexts that may have influenced the differences in their ratings. Clearly, the objective underlying such discussions is to arrive at agreement over the prioritization of the child's intervention goals.

#
#

COMPARISON OF PARENTAL AND PROFESSIONAL JUDGMENTS OF CHILDREN'S SPEECH AND LANGUAGE COMPETENCIES

A study was undertaken to obtain parental judgments of their child's speech and language competencies and to compare parental judgments to those of an SLP. The children of interest were in the 3- to 5-year-old age range, those eligible for services under Part B of the Individuals with Disabilities Education Act (IDEA).

Subjects

The SLAS questionnaire was completed by the mothers and fathers of each child enrolled in LAP. In addition, an SLP completed the survey for all children. Only children with ratings from all three raters were included. Complete sets of SLAS ratings from mothers, fathers, and the SLP were available for 34 children.

All children were between the ages of 3 and 5 years and demonstrated normal intelligence as measured by the Kaufman Assessment Battery for Children (Kaufman & Kaufman, 1983). Fourteen children were developing language normally (ND), whereas 20 children were diagnosed with Specific Language Impairment (SLI) at the time of initial enrollment in LAP. The children in the ND group were developing normally in all aspects of development. The children in the SLI group met standard exclusionary criteria. They all had normal vision and hearing and no known history of neurological or social-emotional disorders. In addition, the children with SLI scored one or more standard deviations below the mean on two or more standardized measures of speech and language development.

The families that participated in this study represented a wide range of socioeconomic backgrounds. The majority of mothers of children in both groups had completed some college or attained a B.A. (ND = 91%, range = high school degree to B.A.; SLI = 78%, range = 9th grade to M.A.). All fathers had completed 12 or more years of education; the majority had also attended some college or attained a B.A. (ND = 73%, range = some college to Ph.D.; SLI = 47%, range = 12th grade to Ph.D.).

#

Procedure

The SLAS was distributed to the parents and the SLP at approximately the same time each semester. All respondents were able to complete a form in approximately 10 minutes. Over a 4-week period of time, the LAP Parent Services Coordinator asked parents to complete the questionnaire at appropriate times after dropping their child off in the preschool or prior to picking up their child. For parents who did not transport their children to the preschool, forms were sent home and they were instructed to complete the forms independently.

The SLP involved in this study was the LAP Educational Coordinator. As the Lead Teacher, she observed these children's communication skills daily. She completed 3 to 4 forms per day over a 2-week period of time embedded within the parent distribution period.

#

Results

Ratings for individual SLAS items were used to compute a mean rating for the five composite SLAS scales (refer to [Table 1]). Composite scale scores were obtained by adding the ratings for the individual items and dividing by the total number of items. Thus, composite scale scores were used in the following analyses.

The first question addressed whether parents and the SLP rated the children's speech and language competencies in a similar manner. Interrater reliability was assessed with zero-order correlations for the orthogonal comparisons (a) mothers vs. fathers and (b) parents vs. the SLP. These correlations indicated that interrater reliability was moderately high to high for all five scales (see [Table 2]). The mean score for each composite scale showed that, as a group, fathers typically rated their child's skills higher than did mothers, who in turn rated their child's skills higher than did the SLP (see [Table 3]). In addition, parental ratings were more restricted than were the SLP's ratings, as evidenced by somewhat smaller standard deviations.

Table 2
Interrater Correlations for Individual SLAS Composite Scales[*]
Scale	Mothers vs. Fathers	Parents vs. SLP
Assertiveness	0.85[*]	0.84[*]
	(33)	(33)
Responsiveness	0.80[*]	0.80[*]
	(33)	(33)
Semantics	0.77[*]	0.85[*]
	(33)	(33)
Syntax	0.90[*] [†]	0.82[*]
	(34)	(34)
Articulation	0.90[*] [†]	0.88[*]
	(34)	(33)

^* Number of children for each correlation is given in parentheses.

^† p < 0.001.

Table 3
Composite Scale Means and Standard Deviations
Scale	Mothers		Fathers		SLP
	Mean	SD	Mean	SD	Mean	SD
Assertiveness	4.13	1.39	4.70	1.51	3.58	1.87
Responsiveness	3.97	1.43	4.39	1.32	3.54	1.89
Semantics	4.13	1.29	4.38	1.40	3.57	1.61
Syntax	3.90	1.63	4.34	1.58	3.47	1.83
Articulation	3.75	1.57	3.91	1.67	3.49	1.79

The second question addressed whether there were certain aspects of speech and language competencies about which parents and professionals were more likely to agree or disagree. We were also interested in whether differences would be influenced by a child's communicative status. Therefore, each composite scale was analyzed using a 2 × 3 mixed ANOVA with repeated measures. The be- tween-group factor was the child's Language Group (ND vs. SLI), whereas the repeated factor was Rater (Mothers vs. Fathers vs. SLP). Given the five individual analyses, a 0.01 alpha level was used to maintain the experiment-wise alpha level at 0.05. These results are presented in [Table 4].

Table 4
F-values for Repeated Measures ANOVAs
	Assertiveness	Responsiveness	Semantics	Syntax	Articulation
Language Group	32.71[*]	33.94[*]	55.24[*]	42.61[*]	37.06[*]
Rater	16.97[*]	7.27[*]	10.65[*]	11.27[*]	1.94
Language Group X Rater	3.53[†]	4.76[‡]	1.14	0.91	0.68

^* p < 0.001.

^‡ p < 0.01.

^† p < 0.05.

Language Group was significantly different for all five scales, confirming the between-group differences that were expected (all F> 32.71, p < 0.001). Main effect Rater differences were found for four scales: Assertiveness (F(2, 62) = 16.97, p < 0.001); Responsiveness (F(2, 62) = 7.27, p < 0.001); Semantics (F(2, 62) = 10.65, p < 0.001); and Syntax (F(2, 64) = 11.27, p < 0.001). Importantly, there were no Rater differences on the Articulation scale. Posthoc Scheffé tests indicated that fathers rated their children significantly higher than did mothers on the Assertiveness scale (F(1, 31) = 14.28, p< 0.001) and on the Syntax scale (F(1, 32) = 11.80, p < 0.01). When averaged together, parents' combined ratings were significantly higher than the SLP's ratings for all four scales (all F > 8.49, p < 0.01). Finally, interaction effects were apparent only for the Responsiveness scale (F(2, 62) = 4.76, p < 0.01). As can be seen in [Table 5], there was close agreement between the parents and the SLP for children in the ND group, but parents of the children with SLI rated their children's responsiveness significantly higher than did the SLP. The same trend was also found for the Assertiveness scale, although this interaction was only marginally significant (F(2, 62) = 3.53,p = 0.04). There were no significant differences between the Raters as a function of the child's Language Group for the remaining scales.

Table 5
Composite Scale Means by Language Group and Rater
Composite Scale	ND Group		SLI Group
	Parents	SLP	Parents	SLP
Assertiveness	5.50 (0.97)	5.17 (1.18)	3.63 (1.11)	2.47 (1.39)
Responsiveness	5.20 (0.93)	5.14 (1.23)	3.51 (0.98)	2.42 (1.40)
Semantics	5.38 (0.88)	4.95 (1.19)	3.39 (0.72)	2.60 (1.05)
Syntax	5.45 (1.08)	5.04 (1.39)	3.19 (1.11)	2.38 (1.20)
Articulation	5.20 (1.02)	5.03 (1.38)	2.87 (1.12)	2.50 (1.24)

In summary, parent and professional ratings of children's speech and language abilities were highly correlated with one another. These strong correlations between parents and the SLP are consistent with previous indirect communication assessment of global language skills by Nass, Watts, Grissom, and Oshrin (1981). In addition, the mean ratings of the parents and the SLP were highly similar for the Articulation scale. This was true for parents of children in both the ND and SLI groups. On the other hand, the four scales on which Rater differences were apparent related to children's language abilities. Overall, Rater differences were most pronounced for the Assertiveness scale. For the parents of children with SLI, differences were most apparent for both conversational assertiveness and responsiveness.

#
#

Clinical Implications

The means in [Table 5] illustrate that the combined parent ratings for every composite SLAS scale were always greater than the SLP ratings for both language groups. This finding is consistent with previous research indicating that parents tend to rate their children's developmental status higher than do professionals (see Sheehan, 1988 for a review). What is most interesting is that differences between parent and SLP judgments varied considerably, depending upon the communicative dimension being evaluated. In the following sections, we focus primarily on the differences found between the parents of children with SLI and the SLP.

Ways in which the raters agree. In this study, the parents and SLP were in strong agreement in their judgments of children's articulation skills and general intelligiblity. As a group, parents of children with SLI rated their children's articulation skills the lowest. In many ways, parents' sensitivity to speech skills is understandable. Limited intelligibility interferes most often with parents' ability to understand their children. In turn, parents may find it difficult, and at times frustrating, to sustain extended verbal interactions with their children. The immediate consequences of limited intelligibility for verbal interactions between parents and children may sensitize parents to this particular dimension of communication skills, and therefore, result in closer agreement with the SLP.

Ways in which the raters disagree. The mean ratings between the parents of children with SLI and the SLP differed by approximately one scale point on SLAS composite scales pertaining to the dimensions of language form, content, and use (range = 0.79 to 1.16). There may be a number of reasons why parent and SLP judgments differed on these dimensions. One possibility is that children's use of language differs between home and school settings and as a function of the interactive partners found in these settings. Children with limited language competency may be more comfortable initiating conversations with parents or siblings at home than when negotiating with peers for favorite toys in a preschool classroom. At the same time, partner or setting differences seem to be an unlikely explanation for differences observed on the Syntax scale. It may be the case that parents are less sensitive to technical aspects of linguistic competency such as the use of determiners, auxiliaries, or inflectional morphemes.

Recommended Uses of the SLAS

In conclusion, the Speech and Language Assessment Scale is a field-tested, reliable, and valid tool that can be used to help parents describe their child's speech and language competencies. Clinicians can use this instrument in a variety of ways. First, they can use it as a way of evaluating the likely validity of formal speech and language assessments. If there are large discrepancies between parent reports and a clinician's findings, the discrepancy bears further examination. Second, the SLAS can be used to facilitate a collaborative relationship between parents and clinicians during IEP planning. Providing parents with an instrument such as the SLAS to complete prior to the IEP conference may enable them to come to the conference with more specific questions and comments about their child's communicative competencies and to participate more fully in the IEP process. Third, the SLAS can be used to review differences that are apparent between parents and clinicians. These discussions can provide important opportunities for parents to inform clinicians about their children's competencies during dinnertime or bedtime routines or when being introduced to new people or in other everyday contexts. In addition, clinicians can use this opportunity to describe their observations of children's verbal interactive skills in the preschool classroom or to educate parents about aspects of semantic or syntactic development that they are particularly concerned about. Fourth, the SLAS could be used as a measure of parent satisfaction with their child's progress in the intervention program. By completing the SLAS at periodic intervals, parents and clinicians can review the ways in which children have made the greatest gains and those in which little progress was noted. Finally, clinicians should recognize that parents differ in their levels of interest in the process of ongoing assessment. Sheehan (1988) recommended that the degree of parental involvement be tailored to the individual parent's expressed interest. The point here is that clinicians should provide parents with a means to increase their involvement, if they so choose, not to burden the parents with just another form to complete. Bearing these thoughts in mind, we believe that the SLAS is a clinically useful tool that can help parents and clinicians work together to provide better services to children with communicative disorders and their families.

#
#

APPENDIX A

LANGUAGE ACQUISITION—PRESCHOOL

Speech and Language Assessment Scale

Rater's Name: Child's Name:

Date: School:

Relationship to Child:

Please rate your child's language and social skills compared to other children his or her own age.

1. My child's ability to ask questions properly is: Comments:

1 2 3 4 5 6 7