As delayed speech is one area where retrospective reports have been utilised, we seek to compare contemporaneous and retrospective parental accounts of concerns over speech difficulties in preschool children. Comparing original accounts of child behaviour to retrospective accounts is the most powerful design for examining the reliability of retrospective recall.

Hypothesis: That for a child treated with speech and language therapy, or an official intervention, a mother would be more likely to recall concerns from the early years, that had not been reported at the time. That if the child has not had S & L therapy or intervention, mothers will be more likely to 'forget' their initial concerns. In general, we expect that mother's recall will be systematically biased by such an intervention.

Method: Secondary analysis to cross compare contemporaneous reports about worries regarding speech development measured when the study children were age between 18 and 30 months, with retrospective reports concerning whether mothers had had worries about children's speech development when in their early years. The retrospectively recalled worries were reported ten years later. We will determine whether the children had ever had any speech therapy or educational intervention and see if this is correlated with accuracy of recall.

Independent variable:SEN provision and intervention (therapy) , Outcome:Difference between contemporaneous and rerospective recall.