Resources Required to Use the Calculator Sheet
Reporting the Results of the Calculators
Effective identification of language impairment among ESL children requires comparing an ESL child's test scores to those of other ESL children. The information in this section was obtained by analyzing test scores from ESL children with and without language impairment. For more information, see Project Overview and Characteristics of ESL Child Participants.
This section describes how to use the Language Development Calculator sheet to estimate the probability that a child has language impairment, based on scores from different combinations of tests. To begin, download the CHESL Language Development Calculator sheet (.XLS).
Each calculator uses a combination of the following scores plus the child's length of exposure to English. Not all test and sub-test scores are needed for every calculator. The minimum is: ALDeQ, TEGI Screener and CTOPP non-word repetition (calculator 6).
The following information will help you use the calculator sheet correctly:
- Calculators give a probability from 0–1.0. Probabilities of 0.5–1.0 indicate that a child's test performance patterns are more consistent with the children in this research who have language impairment.
- The sensitivity and specificity of each combination of test scores is given on the calculator sheet:
- Sensitivity expresses that calculator's ability to correctly classify children as language impaired (expressed as a percentage).
- Specificity expresses that calculator's ability to correctly classify children as typically developing (expressed as a percentage).
The results of the Language Development Calculators are reported differently than individual test scores. This is because:
- The probability score is calculated based on examining a combination of test scores, each with different weightings.
- ESL Children's length of exposure to English is included in the calculation (as opposed to monolingual children, for whom age is equivalent to exposure).
- Test scores are compared ("norm-referenced") to two comparison groups: typically-developing children and language-impaired children.
Here is an example description of how the probability score was obtained and what it means, for reporting purposes:
The child's scores from a group of tests, __________, together with the child's length of exposure to English, were entered into a binary logistic regression formula. Standard scores were entered to take into account the child's age. The formula calculated the probability of that child's test scores patterning like ESL children with typical development as compared to ESL children with language impairment. This calculation is a form of norm-referencing to an ESL comparison group across several tests in combination. The probability score __________ can be interpreted as follows:
Top of Page
A score of 0–.49 indicates a low probability of the child having language impairment.
A score of .50–1.0 indicates a high probability of the child having language impairment.