Relation of SES, Language, and Parent Education in Young Children Who Stutter
CORRIN G. GR AHAM, PH.D., CCC-SLP AND KIA N. HARTFIELD, M.S., CCC-SLP • VANDERBILT UNIVERSITY MEDICAL CENTER
ABSTRACT
The purpose of this study was to examine the relation of socioeconomic status (SES) and parent education on the language abilities of 296 preschool children who do (CWS, n=166) and do not stutter (CWNS, n=130). Participants' performance on normreferenced speech-language measures were correlated to SES, and parental education using the 2-Factor Index of Social Position (Myers & Bean, 1968). Results indicate contradicting correlations for CWS and CWNS with CWS showing a stronger relationship between overall SES status and the influence of parental education on language skills. One possible conclusion is that CWS, when compared to CWNS, are more reactive to and more influenced by environmental factors contributing to SES. Research supported in part by an NIH research grant, (DC00523) to Vanderbilt University. Thus, the purpose of this study was to examine the relation of SES, paternal and maternal education, and the language abilities of preschool children who do and do not stutter.
METHOD
Participants: Participants were 296 (CWNS n=130, CWS n=166) children between the ages of 2;6 and 6;3 who volunteered to take part in a series of studies through the Vanderbilt University Developmental Stuttering Project (VUDSP). Data Collection and Analysis • Speech, language, and hearing screening. As part of a pre-test screening to determine inclusion/exclusion for participation in the VUDSP, each participant was administered a variety of speech-language measures including: the Test of Early Language Development (TELD- version 2 or 3), the Peabody Picture Vocabulary Test (version R or III), and the Expressive Vocabulary Test (EVT), to measure receptive and expressive language, and vocabulary respectively. • The race of each participant was based on parental reports according to the National Institutes of Health (NIH) categories for race (i.e., American Indian or Alaska Native, Asian, Black or African American, Native Hawaiian or Other Pacific Islander, and White) and/or ethnicity (i.e., Hispanic or Latino, and Not Hispanic or Latino). Socio-economic status was determined from parent report of paternal occupation and level of education using the 2-factor index of social position (Myers & Bean, 1968). This index provides scores on a continuum (i.e., 11 to 77) with 11 indicating the highest possible level of occupational status (e.g., CEO) and education (e.g., M.B.A.) and 77 indicating the lowest level of occupational status (e.g., dishwasher) and educational level (e.g., less than seventh grade education). All participants also participated in a parent-child interaction to permit analysis of disfluencies/stutterings (e.g., Logan & Conture, 1997) for determination of talker group (e.g., CWNS or CWS). Group means and/or percentages by talker group were calculated for the following variables: age, gender, race, SES 2-factor score for the father, father's education, mother's education. Analysis of variance (ANOVA) was used to calculate differences between groups where it was appropriate (e.g., chronological age, 2-factor SES score, father's education, mother's education, and performance on language measures). Visual inspection of percentage data was used where inferential statistics were not appropriate (e.g., gender and race). Pearson product-moment correlations were used to detect relations between SES and language measures.
BACKGROUND
Previous empirical studies examining the language skills of children from varying socio-economic backgrounds have indicated significant differences between economically disadvantaged children and their more economically advantaged counterparts (Deutsch, 1965, 1963; John, 1963; Jordan & Robinson, 1972; Larson & Hermann, 1974; Milner, 1951; Osser, Wang, & Zaid, 1969; Robinson, 1965; Smith & Lefrancois, 1972; Thomaneck, 1972). Overall, language development appears to develop at a faster rate for high SES children, which is most apparent at twoyears of age. As a result, there has been an increasing trend in research involving young children who stutter (CWS) to specify and report any relative differences in participants' socioeconomic status (SES) as well as racial/ethnic differences when reporting language related findings (Anderson, Pellowski, & Conture, 2005; Anderson, Pellowski, Conture, & Kelly, 2003; Byrd, Conture, & Ohde, 2006; Yairi, & Ambrose, 1999). Furthermore, research suggests that CWS with low SES are excluded from stuttering research at a significantly higher rate than CWS with high SES, a finding not present in children who do not stutter (CWNS; Graham, Conture, & Zackheim, 2003). Therefore, it is not unreasonable to suggest that their may be a significant relationship between the SES and language abilities for CWS, but not for CWNS. Perhaps the influence of SES on language development in CWS is a result of subtle to not so subtle differences in language skills coupled with CWS tendency to be highly reactive to environmental factors (Anderson,Pellowski, Conture, & Kelly, 2003). The present study considers the SES of both CWS and CWNS in relation to their standard scores on norm-referenced tests of receptive and expressive vocabulary as well as receptive and expressive language abilities. The Hollingshead Two Factor Index of Social Position (Myers & Bean, 1968) was used to calculate SES by yielding a composite score for the head of household (typically the father) based on education and occupation.
RESULTS
Descriptive data:
GENDER AND RACIAL BREAKDOWN OF PARTICIPANTS BY TALKER GROUP
100 90 80 70 60 50 40 30 20 10 0
(TELD-R% r=-0.192, p<0.018; TELD-Exp% r= -0.266, p<0.002; TELD-SLQ% r=-0.240, p<0.004; EVT % r=-0.245, p<.004; PPVT % r=-0.303, p<0.001). However, for CWNS, no significant correlations to the language measures were found. Correlation between Paternal Education and Language 2. For CWS, paternal education indicated significant correlations to all language measures (TELD-Rec% r=-0.270, p<0.002; TELD-Exp% r= -0.312, p<0.001; TELD-SLQ% r=-0.316, p<0.001; EVT % r=-0.322, p<.001; PPVT % r=-0.396, p<0.001). However, for CWNS, a significant correlation was found for the EVT only (r=-0.192, p<0.036). Correlation between Maternal Education and Language 3. For CWS, maternal education indicated significant correlations to percentile ranks for the TELD-Expressive (r=-0.150, p<0.054), EVT (r=-0.223, p<0.008), and PPVT (r=-0.193, p<0.018). For CWNS, a significant correlation was found for all language measures (TELD-Rec% r=-0.171, p<0.052; TELDExp% r= -0.208, p<0.023; TELD-SLQ% r=-0.215, p<0.020; EVT % r=-0.391, p<.001; PPVT % r=-0.267, p<0.005).
Regression Summary Table
Father’s 2-factor-Score Father’s Education EVT Mother’s Education TELD-R, TELD-E, TELD-SLQ, EVT, PPVT TELD-E, EVT, PPVT
Percent
CWS CWNS
Ma
le
Fe
ma
le C c au
as
ian
Af
an ric
er Am
n ica
ian As
sp Hi
an
ic M
a i-r ult
c ia
l
Results of ANOVA indicated no statistically significant differences between the CWS and CWNS for the following variables. (All error bars represent the standard error of measure): 1. Chronological age in months by Talker group, F(1,294) = 0.38, p < 0.618. 2. Hollingshead 2-Factor Index Score (based on father) by Talker group, F(1,292) = 0.250, p < 0.618.
FATHER'S 2-FACTOR-SCORE
CHRONOLOGICAL-AGE (MONTHS)
60 50 40 30 20 10 0 31 26 21 16 11
CWNS
No significant correlations TELD-R, TELD-E, TELD-SLQ, EVT, PPVT
3 2.5 2 1.5 1
CWS
TELD-R, TELD-E, TELD-SLQ, EVT, PPVT
DISCUSSION
When 0.5
CWS CWNS
CWS
CWNS
3. Paternal Education by Talker group, F(1,283) = 0.152, p < 0.697. 4. Maternal Education by Talker group F(1,287) = 0.509, CWS p < 0.476. CWNS
PARENT EDUCATION BY TALKER GROUP
3 2.5 2 1.5 1 0.5 0
Father Education
Mother Education
A linear regression was calculated to determine if the Hollingshead 2-factor-score, paternal education, and maternal education scores were correlated with the percentile ranks from the Test of Early Language Development subtests of receptive language (TELD-R), expressive language (TELD-E), and spoken language (TELD-SLQ). A linear regression was also calculated for the percentile ranks on the Expressive Vocabulary Test (EVT) and the Peabody Picture Vocabulary Test (PPVT; See Regression Summary Table for results). Correlation between Hollingshead 2-Factor Index and Language 1. For CWS, the Hollingshead 2-Factor Index (based on father) indicated significant correlations to all language measures
compared to CWNS, it appears that for CWS the following factors of 1) SES, 2) paternal education, and 3) maternal educa0 tion -Father Education separately - have a greater impact on their overall speechMother Education language abilities. However, for CWNS it appears that the only factor that seems to have an impact on their overall speech-language abilities is maternal education. These findings may suggest that CWS are more reactive to environmental factors (i.e., maternal and paternal education and occupation) that reportedly, influence speech-language development. More specifically, it appears to be the case that the speech-language abilities of CWS are more influenced by the education and interaction styles of both parents more so than CWNS. These findings may suggest that as we continue to explore all of the variables that are thought to contribute to the behavior of stuttering, consideration should be given to the influences of the socio-educational environment as well as temperamental variables (i.e., reactivity and regulation) of individuals who stutter. The more we know about how factors, other than the actual behavior of stuttering itself, contribute to and/or exacerbate stuttering, the more refined our understanding, identification, and treatment of people who stutter will become.
Presented to the annual convention of the American Speech-Language-Hearing Association, Miami, FL (November, 2006). Poster constructed by the Graphics Core of the Vanderbilt Kennedy Center, Vanderbilt University, supported in part by NICHD Grant P30 HD15052. The Vanderbilt Kennedy Center is devoted to unlocking the mysteries of the brain and child development, and improving the lives of people with disabilities. kc.vanderbilt.edu