Listed below is a glossary of the psychometric terms we used to define our selection
criteria for recommending highly validated screening tools within their respective
Selection Criteria |
Recommended Tools |
A standardized screening tool has been tested on hundreds of children from diverse
backgrounds geographically and demographically. This ensures that the tool represents
the entire population of the U.S., and internationally, in certain cases. Within
this category, supporting information that describes the test population, the
length of time on the market, and industry peers who have reviewed the results
Inter-rater reliability agreement demonstrates consistent results between different
examiners. Test-retest reliability shows reliability between tests conducted
over short periods of time. Internal consistency exists among test items within
a tool to prevent redundancy.
Although there are many types of validity, a screening tool recommended by First
Signs should test for content validity and concurrent validity. Content validity
tests a developmentally appropriate range of skills. Concurrent validity shows
how well a screening tool compares to other similar measures on the market.
Accuracy or Criterion-based Validity
regards accuracy as the most critical category for choosing a screening
tool, particularly in the areas of sensitivity and specificity. Sensitivity and
specificity provide the bottom line measures on whether a tool is highly validated
or not. Sensitivity illustrates how often a screening tool accurately identifies
children suspected to be at risk for a developmental disorder, such as autism.
Specificity illustrates how often a screening tool accurately rules out children
not at risk (i.e., those who meet typical developmental milestones).
For a screening tool to receive a recommendation from First Signs, both sensitivity
and specificity percentages must exceed 70 to 80%. Ideally, a tool should show
sensitivity and specificity in the 90th percentile, though no screens for developmental
disorders or autism are available currently in this high percentage range. Additionally,
sensitivity should be equal to or greater than specificity. Sacrificing sensitivity
for a higher specificity is not in line with our goals, even though higher specificity
keeps costs of over-referral down. It is critical that as many children as possible
who present developmental delays are screened and referred for diagnosis and
intervention. High sensitivity will accomplish this.
Another meaningful accuracy criterion is positive predictive value, which indicates
the percentage of children who fail the screen and then receive a confirming
Developmental screening tools must follow appropriate developmental milestones
to receive a positive rating from First Signs. Developmental milestones provide
important guidelines for tracking healthy development in early childhood.
The DSM-IV is the abbreviated term for the Diagnostic and Statistical Manual
of Mental Disorders - Fourth Edition (DSM-IV), published by the American Psychiatric
Association, Washington D.C., 1994, which is the main diagnostic reference of
mental health professionals in the U.S.
requires that a screening tool meets key diagnostic criteria from
the DSM-IV for the following disorders:
299.00 Autistic Disorder
299.80 Asperger’s Disorder
Our current focus is developmental and
behavioral disorders. Please revisit
this Web section in the future as we expand our list of disorders and related
To view DSM-IV criteria for other Autism Spectrum Disorders (ASDs), also known
Pervasive Developmental Disorders (PDDs), click here.
Diagnostic Criteria for 299.00 Autistic Disorder, Diagnostic
Manual of Mental Disorders (4th ed., pp. 70-71) Washington, DC: American Psychiatric