This study investigates the consequences of using different methods when making conclusions about the effect of IFRS. The study's contribution is related to the suggestion that research design must be considered when analysing the results of prior studies on the adoption of IFRS. In summary, our results demonstrate that the impact of IFRS on accounting quality measured as financial analysts' information environment depends on the model used. Using the same model as in prior research (the BKLS model) our results confirm earlier studies that IFRS adoption improves the information environment for analysts. However, when using a more sophisticated model no such improvement can be found.