This paper builds on feminist elaborations of Bernstein's code theory to engage in a series of thought experiments with interview data produced during a co-inquiry design-based research intervention project. It presents three accounts of thinking/writing with data. Our purpose in presenting three different accounts of interview data is to demonstrate the relation between theory and empirical data. In the first two accounts, interview data are interpreted and performed through the lens of theory. By contrast, in the third account attention is paid to the ways in which care is practised not only in terms of policy enactment, but also research enactment. Empirical data are not moulded to fit generalisable theoretical frameworks. Rather, empirical data push back on theoretical concepts in a collaborative thought experiment.