Most existing research on early identification of learning difficulties has
examined the validity of methods for predicting future academic problems.
The present study focused instead on the sensitivity of kindergarten teache
rs to learning problems in their students and on the continuity of teacher-
identified problems over time. To identify early learning problems, kinderg
arten teachers in a suburban school district rated student progress toward
six academic objectives as satisfactory or unsatisfactory. Twenty percent o
f the district's 303 kindergarten children received unsatisfactory ratings
in at least one area. Thirty-eight of these children (identified group) wer
e matched to 34 children with satisfactory ratings in all areas (nonidentif
ied group). Results of testing conducted during kindergarten revealed poore
r academic achievement in identified children than in nonidentified childre
n. Children from the identified group also performed more poorly than child
ren from the nonidentified group on tests of phonological processing and wo
rking memory/executive function and were rated by teachers as having more b
ehavior and attention problems and lower social competence. Follow-up of th
e sample to first grade documented continued learning problems in the ident
ified group. These findings support the use of teacher judgements in early
detection of learning problems and argue against reliance on discrepancy cr
iteria.