A parent-report gender identity questionnaire for children

L. Johnson, S. Bradley, A. Birkenfeld-Adams, M.Radzins Kuksis, D. Maing, J. Mitchell

This paper reports on the psychometric properties of a 16-item parent-report Gender Identity Questionnaire, originally developed by P. H. Elizabeth and R. Green (1984), to aid in the assessment of children with potential problems in their gender identity development. The questionnaire, which covered aspects of the core phenomenology of gender identity disorder (GID), was completed by parents of gender-referred children (N = 325) and controls (siblings, clinic-referred, and nonreferred; N = 504), who ranged in age from 2.5_12 years (mean age, 7.6 years). Factor-analysis indicated that a one-factor solution, containing 14 of the 16 items with factor loadings =.30, best fit the data, accounting for 43.7% of the variance. The gender-referred children had a significantly more deviant total score than did the controls, with a large effect size of 3.70. The GIQ total score had negligible age effects, indicating that the questionnaire has utility for assessing change over time. The gender-referred children who met the complete DSM criteria for GID had a significantly more deviant total score than did the children who were subthreshold for GID, although the latter group had a mean score that was closer to the threshold cases than to the controls. With a specificity rate set at 95% for the controls, the sensitivity rate for the probands was 86.8%. It is concluded that this parent-report gender identity questionnaire has excellent psychometric properties and can serve as a useful screening device for front-line clinicians, for whom more extensive, expensive, and time-consuming assessment procedures may be precluded.