This paper proposes the principal features of a research aiming to investigate gender stereotypes from a perspective that involves the past, the present and the future of the relations between the sexes. In particular we focus on the construction of an instrument for the measurement of attitude towards female stereotypes. 5-point Likert scales was used to determine attitude towards stereotypes associated to six different dimensions and an online survey was conducted in order to obtain responses from different countries. The collected data have been exploited for Likert item selection through the implementation of the Partial Credit Model, so that the selected items are Rasch-homogeneous and comply with the restrictive Rasch criteria