Characteristics of the DYMOND Normative Sample
Number of children tested: n = 1454. The DYMOND normative sample, which represents children who have no known disorders that might affect language learning potential, decoding learning potential, or underlying cognitive-linguistic factors, provides the examiner with a “normal” comparison group that can be used as a reference to which children administered the DYMOND can be compared, facilitating the detection of divergence from the norm.
Data collection window: 2018-2024
Age/grade: Kindergarten – eighth grade, ages 5;0-14;5 (years;months). Six-month intervals were specifically included for ages 5;0-8;11 based on the rapid rate of language development in younger children and based on statistical analyses that indicated significant differences between six-month age bands in the mean performance of subtests sensitive to language development. One-year intervals were included for children ages 9;0 to 9;11 and for children ages 10:0 to 10;11 based on statistical analyses. Ages 11;0 to 14;5 were grouped together because statistical analyses indicated that was no significant difference between those ages.
Number of sites/states: 80 testing sites over 36 states across the U.S. In several states all children in entire classrooms or schools were individually administered the DYMOND.
Number of multilingual children included in normative sample: n = 359 (24.7%). Multilingual children were included in the normative sample unless they could not understand very simple English directions necessary to perform the assessment tasks. For example, children who received a score of 1 on the WIDA assessment (WIDA, 2020) were still eligible to be included in the normative sample as long as their MLU was ~2.0 and they could understand the examiner’s instructions. This resulted in a very diverse group of typically developing kindergarten through eighth grade students in the normative sample which more closely reflects the range of English language proficiency and multilingualism found in the U.S. (>20%).
Reference Standard for Language and Decoding Learning Potential
Prior to the administration of the DYMOND, it was necessary to ascertain the factual state of each child’s language or decoding learning ability, which was dichotomized into the classification of disorder or no disorder. This classification from the reference standard was used to obtain sensitivity and specificity data for the DYMOND. There were 412 children with language and/or decoding disorder, 295 children had language disorder and 162 had decoding disorder. Of those children with language disorder, 67 (22.7%) also had a decoding disorder.
The reference standard for language disorder (n = 295) was the diagnosis by a speech-language pathologist (SLP) of developmental language disorder, a language disorder due to a concomitant disorder, or eligibility for special education services for speech or language impairment with language goals. Children who had been referred for a language evaluation but who had not yet been determined to be eligible for special education services or did not yet have an IEP in place were only included in the language disorder group if an SLP confirmed that the child had a language disorder.
The reference standard for a decoding learning disorder (n = 162) was the diagnosis of a reading disorder with specific difficulty with decoding by a special educator, reading specialist, school psychologist, or any other qualified professional trained to make such diagnoses, or eligibility determination for special education services for specific learning disability or any other of the 13 IDEA categories with specified decoding-related difficulty. We held the kindergarten students to a more stringent reference standard, where an IEP with specified concerns for decoding learning ability had to be in place in order for a child to be diagnosed with a decoding learning disorder.
