Revising the measurement of Eriksonian ego identity in contemporary Italian adolescents: the EPSI-IT

Introduction: The Identity subscale of the Erikson Psychosocial Stage Inventory (EPSI) is the most widely used brief measure of Eriksonian ego identity, but it carries a historically situated gender item and has no published Italian validation. Instruments inherit the assumptions of the era that produced them, and an item can lose validity as those assumptions shift. Methods: In two waves of an Italian school cohort in Rome (2025, N = 3,068; 2026, N = 2,896), we examined the original EPSI and developed the revised EPSI-IT. Results: Among the models compared, a bifactor solution fit best in both waves (EPSI-IT ωH = 0.64, ECV = 0.54), though absolute fit was moderate; the structure supports a total score complemented by a synthesis subscale, whereas clarity (reversekeyed confusion) adds little beyond the general factor. The 11-item EPSI-IT showed a coherent structure and a theory-consistent nomological network, and the harmonized core was invariant across groups and waves in terms of the loading level, with a minor intercept-level decrement. Across-wave criterion attenuation appeared equally on the constant harmonized items, arguing against an explanation specific to the revision. The gender item did not function as an ego-identity indicator and was removed. Discussion: We recommend the total score as the primary index, complemented by the synthesis subscale, and report preliminary grade-specific reference values for research use. More broadly, a historically situated item—here gender-related certainty used as a marker of ego identity—can come to function as separable content in a contemporary cohort.