Predictive validity of the Fagan Test of Infant Intelligence

Abstract

The participants were 78 healthy infants aged 27, 29, 39 or 52 weeks who were tested for visual recognition memory (novelty preference) with the Fagan Test of Infant Intelligence (FTII) and on the Bayley Scales of Infant Development (BSID). Internal consistency of novelty preference from FTII was poor and test‐retest reliabilities between 27/29 weeks, 29/39 weeks and 39/52 weeks were low and non‐significant. Infants were followed up 1 year later and again at 2 years of age. Correlations between BSID and FTII scores were generally low and non‐significant and, with one exception, FTII did not predict later mental outcome. The FTII score at 52 weeks correlated significantly with the Mental Development Index from BSID (1993 edition) at 2 years of age but not with Stanford‐Binet scores. This correlation coefficient indicated that the Fagan test might account for up to 24% of the shared variance in later outcomes. Nonetheless, the present study suggested that the clinical utility of the current FTII, as a screening device with infants at risk for slower intellectual development, is limited, because of low sensitivity, which did not exceed 37.5%.