This study was motivated by the observation that despite the broad equivalence between sexes on diagnostic criteria for DLD, boys are overrepresented in language intervention services by a factor of more than two to one compared to girls. Consequently, we ask the following: with regard to language disorders…
RQ1: Do girls have fewer needs than boys?
RQ2: Do girls have different needs than boys?
RQ3: Do girls have later needs (i.e. later emergence of disability) than boys?
Hypotheses
======================================================
# H1: Fewer Needs
======================================================
If service delivery is determined by the extent of disability rather than sex bias, the effect of sex on service delivery will weaken when controlling for reading achievement scores.
Rationale: Girls and boys may differ in temperament in ways that serve to protect girls with language impairment from disability, or to increase the risk of disability among boys with language impairment.
======================================================
# H2: Different Needs
======================================================
If the service gap reflects different service delivery paths, then girls will be more likely to be served under SLD than LI relative to boys.
Rationale: This hypothesis will explore the possibility that girls with DLD require services from professionals other than an SLP because their disabilities are largely academic.
======================================================
H3: Later Needs
======================================================
If language disability emerges later in life for girls compared to boys, there will be a significant positive effect of sex on age of service delivery, such that girls will be diagnosed at later ages compared to boys.
Rationale: Girls with DLD may be more likely than boys to meet social expectations in early life. Because of this, we reason that the needs of girls might tend to emerge only when academic demands increase in later years.
Data Description
Datasets used
This project makes use of the Florida Progress Monitoring and Reporting Network (PMRN) database. This is a longitudinal dataset with data collected between 2003 and 2014. Respondents were grade-school aged students in the Florida public school system.
Access date
The version of the PMRN dataset which will be used for statistical analysis was accessed on January 14, 2025 (2025-01-14).
Data collection procedures
This study will uses Progress Monitoring and Reporting Network (PMRN) database, which includes information about students’ demographics, screening and diagnostic measures, and academic achievement. These data are maintained by teachers and school administrators across the state of Florida. Authorized users are able to enter scores for students for each academic year using a variety of standardized measures.
The data were collected as part of academic progress monitoring from over 1.6 million children in grades 1 to 12 (roughly ages 6 to 18). The data comprise two sets. There are 393,300 children in set one, 48.29% female and 51.69% male. The race and ethnicity composition is 34.54% White, 34.57% Black, 24.80% Hispanic, 1.44% Asian, 0.31% Native American/Pacific Islander, and 3.92% Mixed. There are 1,287,728 children in set two, 48.02% female and 51.94% male. The race and ethnicity composition is 37.64% White, 25.59% Black, 24.00% Hispanic, 1.96% Asian, 0.37% Native American/Pacific Islander, and 2.75% Mixed. The racial and ethnic composition of the sample reflects the demographics of the state of Florida in the years 2002-2014. The size and diversity of the sample and its longitudinal nature make it ideal for addressing the goal of determining the extent to which disability (rightly) drives service delivery and the extent to which sex predicts service receipt.
Access to the original PMRN dataset is restricted due to U.S. privacy regulations and cannot be publicly shared. However, we have created a de-identified version of the datasets used in our analyses, which we have shared in the data repository. We have also included code for reproducing these restricted datasets from the original PMRN files.
For more information, please see the original preregistration for this project on OSF:
https://osf.io/sb5qr/overview?view_only=7aa7170cba384ef3ad05504c4d0291fc
