“…The already mentioned mixed-methods approach, discovery-oriented task analysis (Greenberg, 2007), might be very useful for this purpose, because it combines the elaboration of a clinically informed theoretical model with an empirical one in a rational-empirical model. Moreover, also more traditional qualitative methods might be used to explore and inductively build models of the therapeutic process as dynamic open complex system; this could be the case, for example, for grounded theory analysis or phenomenological analysis when we are interested in clients’ and/or therapists’ personal and subjective experiences about the treatment (e.g., Murray, 2002), or for conversational analysis when we are interested in how clients and therapists actually shape the dynamics of their relationships by means of the coordination (i.e., synchronization) of their turn at speaking (e.g., Muntigl & Horvath, 2014). Finally, efforts should be made to explicitly frame within a DS approach – in terms of both hypothesis generation and result interpretation – the studies conducted following specific longitudinal designs (e.g., the microanalytic sequential process design and designs using time-series analysis) or the significant events approach.…”