This article presents preliminary analysis of a test item in a large-scale study design to promote the development of geometric reasoning progression. Two sets of data were analysed to validate the item designed to assess secondary school students' knowledge of a rectangle. The first data set involved 155 Year 4-10 students from seven trial schools across social strata. The second data set involved 585 Year 7-10 students from eleven project schools situated in lower socio-economic areas. The aim was to audit Australian students' knowledge of hierarchy of shapes and document the process of validating a test item. The findings indicated that an iterative process of design, test and redesign, incorporating Sfard's mathematical discourse framework and a multi-stage Rasch analysis, is vital in validating the results. A distinct change in students' reasoning about rectangle is observed and this is not due to age. Moreover, Rasch analysis identified eight distinct thinking zones to assist in mapping out a learning progression for developing geometric reasoning.