In 1974 Sandra Bem Asked 100 American students to select characteristics that they believed to be masculine, feminine, or neutral, then produced a self report questionnaire where people would rate the trait on a scale of 1-7 on how much they identified with it. This was then put into a grid to show how masculine/feminine people were and people who were both were classes as androgynous.
One strength of this study is its high external reliability as it gave consistent results every time it was repeated. Bem showed this through the testretest method, asking the same group of participants to repeat the test 4 weeks later. Furthermore, it has high internalreliability, as Bem split the test into two half's and asked the group of participants to complete it, finding that the results from the separate half's were similar. Finally, the 100 students used by Bem to create the BSRI were evenly split into male and female, and only traits chosen by both men and women were used, meaning that the questionnaire is unlikely to have any genderbias.
this study has been criticised for potentially lacking temporal validity as the characteristics were chosen in the 1970s, and we know that sex role stereotypes change over time. Additionally, the 1-7 scale is subjective so it may lack internal validity, questioning wether the results of the study can be used at all.