Remaining open to quantitative, qualitative and mixed-method designs : an unscientific compromise, or good research practice?