SES Influences on Preschoolers’ Performance on the Preschool
Language Scale and the Peabody Picture Vocabulary Test
Marvin W. Lee Tennessee State University Kathryn Guillot, Elizabeth Spencer, Karen Barako Arndt, Julie Rosenthal, Anna Lineback,
Jamie Fisher, Krystal Werfel, Christina Foran, C. Melanie Schuele Vanderbilt University Alyson Abel University of Kansas
Norm-referenced instruments are widely used in evaluating the language skills of
preschool children to make diagnostic decisions. In this study, we examined the
performance of a group of children from primarily college-educated families from
Nashville, TN on the PLS-4 and the PPVT-III. This group performed substantially
above the test mean. In previous studies of disadvantaged preschoolers from
Nashville, Qi, Kaiser, and colleagues reported that the group performed substantially
below the test mean on the PLS-3 and PPVT-III. Implications for identification of
language impairment and enhancement of language skills are considered.
Norm-referenced instruments are widely used to make diagnostic
decisions regarding the language skills of preschool children. When
assessing the appropriateness of a particular instrument, clinicians
are often advised to consider whether the normative sample included
a diverse representation of children (e.g., mirrored the US census).
Hutchinson (1996) pointed out the limitations of this view, noting that
diverse representation may be insufficient if subsamples of children
perform quite differently on an instrument (e.g., differing distributions
for subsamples of the population).
Although there is substantial evidence in the research literature that
children from economically and educationally disadvantaged families
perform below the test mean (i.e., 100), to our knowledge, no
commercially available language measures report subgroup norms.
There is substantial evidence that the group mean for children from
families of low socioeconomic status (SES) is below age expectations.
In a review of the language abilities of low SES children, Whitehurst
(1997) reported that the mean performance of these children was more
than one standard deviation below children from higher SES families
on measures of receptive vocabulary, expressive vocabulary,
metalinguistic skills, narrative skills, and sentence complexity. To
illustrate, Lonigan et al. (1998) reported the mean on receptive
language for low SES children to be 79.09 (SD = 17.66) and for middle
SES children to be 101.11 (SD = 14.25), d = 1.37.
With respect to the Peabody Picture Vocabulary Test and the
Preschool Language Scale, commonly used instruments in preschool
language assessment, Qi, Kaiser, and colleagues reported group
means for low SES preschoolers to be substantially below the test
means (M =100, SD = 15). For the PPVT-III, Qi, Kaiser et al. (2006)
reported the mean for a African American low SES sample (n = 482) to
be approximately 1.5 SD below the test mean. A similar discrepancy
was noted in a smaller group of European American low SES
preschoolers (n = 52). For the PLS-III, Qi, Kaiser et al. (2003) reported
that the mean for the African American low SES sample (n = 701) was
approximately 1 SD below the test mean. Importantly in both studies,
the distribution of scores approximated a normal distribution. These
findings suggest a need to explore subgroup means on normreferenced measures of oral language ability.
The purpose of this study was to consider the impact of
socioeconomic status, as indexed by maternal education, on
children's performance on the Preschool Language Scale and the
Peabody Picture Vocabulary Test.
On the Peabody Picture Vocabulary Test-III, the group mean was
109.95 (SD = 13.4). On the Preschool Language Scale-4, the
group mean was 116.76 (SD = 13.4). Thus, a group of children
with mothers’ who had at least a four-year college degree
performed substantially better than what is defined as
“average”performance for the population. Only XX% on the
PPVT-III and XX% on the PLS-4 received a standard score below
100. The distribution of scores on the PLS and PPVT
approximated a normal distribution, indicating that the measures
were effective in differentiating strong from weak language skills.
Comparison of Groups Based on Maternal Education
PLS-3 or 4
Low Maternal Education
(Qi et al., 2003; Qi et al., 2006)
High Maternal Education
Comparison of the data for low SES children, reported by Qi et al.
(2003, 2006), to the data reported here revealed discrepant means
and distributions for subgroups defined by maternal education.
However, the distribution of scores, as illustrated by the figures
below, indicated that for the low SES group the group scores
approximated a normal distribution. Thus, as with the higher SES
group, the PLS and PPVT effectively differentiated children with
strong language ability from children with weak language ability.
Qi et al. PPVT-III Distribution
In this study, a comparison of the performance of higher SES
children to the performance of lower SES children indicated that
the range of scores was not comparable. In fact, there was
minimal overlap in group scores. The published test population
norms do not adequately characterize either of these subgroups.
Language impairment is often operationalized as the "low end of
normal" with a cut-off, for example, of one standard deviation
below the age mean. However, with a cut-off of one standard
deviation below them mean, few children from high SES families
but a large proportion of children from low SES families would be
identified as language impaired.
The figures below illustrate the test population norms alongside
the estimated distribution of scores of children from collegeeducated families contrasted with the distribution of scores of
children from economically and educationally disadvantaged
families. These distributions illustrate the challenges of using
these measures diagnose language impairment. The findings in
the current study, along with the findings reported by Qi, Kaiser,
and colleagues (2003, 2006), argue strongly for test publishers to
not only report population distributions, but subpopulation
distributions as well. With this information, clinicians would be
able to make more informed diagnostic decisions.
In the diagnosis of language impairment, the goal is to identify
those children who have difficulty learning language. We could
argue that children who score on the low end of the population
distribution have failed to benefit from language learning
opportunities as compared to same age peers. However, this
argument is only valid if we are comparing a child to a peer
group that has had similar language learning opportunities.
Further Hutchinson (1996) explained application of the test
distribution is only appropriate to the extent that subgroups of
children perform similarly on the measure (i.e., similar
distribution). From the data reported here we argue that if the low
end of normal is a valid approach to identification of language
impairment, then, at least on the PPVT and the PLS, it is
necessary to compare a child to the distribution for the subgroup
of which he or she is a member. Comparison to the test
population distribution is quite problematic.
Qi et al. PLS-3 Distribution
Qi et al., 2006
This study is a secondary analysis of data from children who
participated in a study of word learning (n = 49; Abel & Schuele,
2007) or in a study of complex syntax development (n = 100;
Schuele, 2006). The PPVT-III was administered in both studies; the
PLS-4 was administered only in the complex syntax study. Item 66
on the PLS-4 was not administered.
In the groups of low SES children described by Qi et al. (2003,
2006), XX% of the children received a standard score below 100
on the PPVT-III and XX% on the PLS-3.
Qi et al., 2003
Who is a child with a language impairment?
A child with a language impairment is a child who has difficulty
learning language. A language impairment is often operationalized as
the low end of normal, defined relative to a child’s same-age peers.
This definition assumes that the child has had similar language
learning experiences as his same-age peers. When language measures
correlate with SES, measurement may tap into differences in
experience making it difficult to obtain a true or accurate reflection of
the child’s language learning abilities. To improve accuracy, diagnostic
decisions must be made by comparing children to a peer group
matched for age and SES, resulting in a similar proportion of children
within each SES group identified as language impaired.
Who is a child that does not have strong language skills?
As illustrated, some children have scores that fall outside the range of
average as defined by the test norms but who score within the average
range (+1 SD) for their SES peer group. This group of children may be
viewed as children who do not have strong language skills rather than
children with language impairment. Children who do not have strong
language skills may have limited language learning experiences.
IMPLICATIONS FOR CLINICAL PRACTICE
Policy and practice have often placed enrichment and intervention in
the same category (i.e., Dickinson & Caswell, 2007). However, it is
critical that we differentiate between the purposes, service delivery
method, and children served by enrichment versus intervention.
Children who do not have strong language skills need
• Language enrichment increases the support for language
and facilitates the acquisition of general language
• Language enrichment is provided through enhanced experiences in
the classroom and increasing support for parent-child interaction.
• SLPs serve in a collaborative/consultative role in language
Children who have language impairments need
Language intervention targets the improvement of communication
abilities and function in children with language impairments (ASHA,
• Language intervention is provided through direct and indirect
targeting the acquisition of language skills specific to an
individual child’s needs.
• SLPs serve a primary role in providing language intervention, in
collaboration with other members of the educational team.
American Speech-Language-Hearing Association. (2004). Preferred practice patterns for the profession of speechlanguage pathology [Preferred Practice Patterns]. Available from www.asha.org/policy.
Dickinson, D. K., & Caswell, L. (2007). Building support for language and early literacy in preschool classrooms
through in-service professional development: Effects of the literacy environment Enrichment Program (LEEP). Early
Childhood Research Quarterly. 22, 243-260.
Hutchinson, T. A. (1996). What to look for in the technical manual: Twenty questions for users. Language, Speech, and
hearing Services in Schools, 27, 109-121.
Lonigan, C. J., Burgess, S. R., Anthony, J. L., & Barker, T. A. (1998). Development of phonological sensitivity in 2- to 5year-old children. Journal of Educational Psychology, 90, 294-311.
Qi, C. H., Kaiser, A. P., Milan, S., & Hancock, T. (2006). Language performance of low-income African American and
European American preschool children on the PPVT-III. Language, Speech, and hearing Services in Schools, 37, 5-16.
Qi, C. H., Kaiser, A.P., Milan, S.E, Yzquierdo, Z., & Hancock, T. B. (2003). The performance of low-income, African
American children on the Preschool Language Scale-3. Journal of Speech, Language, and Hearing Research, 46, 576-590.
Whitehurst, G. J. (1997). Language processes in context: Language learning in children reared in poverty. In L. B.
Adamson & M. A. Romski (Ed.), Communication and language acquisition: discoveries from atypical development (pp.
233-266). Baltimore: Paul H. Brookes.
IMPLICATIONS FOR IDENTIFICATION OF LANGUAGE
Ten examiners administered the PPVT-III and PLS-4. Participants
were recruited from six Nashville community preschools. The study
sample included 149 preschool children (78 girls; 71 boys). Mean age
was 52.89 months (SD = 8.6 months; Range = 36 to 67 months). All
children were monolingual English speakers. No child was enrolled
in speech-language therapy. Nearly all children (XX%) were
Caucasian. Of the 82 families reporting maternal education, 97% of
mothers had at least a bachelor’s degree.
This study was supported by NIH/NIDCD RO3 (PI: Schuele)
Contact: [email protected]