Table 2:

Scientific questions and model parameters for a hybrid factorial-SMART

Scientific question		Type	Model (1) parameters
A	Is it beneficial (in terms of month 6 weight loss) to offer (or not to offer) the App at the beginning of the program?	Main effect of initial factor, averaging over the other factors and response status	2θ1
B	Is it beneficial (in terms of month 6 weight loss) to offer (or not to offer) Coaching at the beginning of the program?	Main effect of initial factor, averaging over the other factors and response status	2θ2
C	Is it beneficial (in terms of month 6 weight loss) to offer (or not to offer) Meal at week 2 for non-responders?	Main effect of subsequent factor for non-responders, averaging over the other factors and response status	2θ4
D	Is the App more beneficial (in terms of month 6 weight loss) with or without Coaching?	Two-way interaction between the two initial factors, averaging over the subsequent factor and response status	2θ3
E	Is the App more beneficial (in terms of month 6 weight loss) with or without offering Meal to non-responders at week 2?	Two-way interaction between initial factor and subsequent factor for non-responders, averaging over the other initial factor and response status	2θ5
F	Is Coaching more beneficial (in terms of month 6 weight loss) with or without offering Meal to non-responders at week 2?	Two-way interaction between initial factor and subsequent factor for non-responders, averaging over the other initial factor and response status	2θ6
G	Does the interaction between App and Coaching vary depending on whether or not Meal is offered to non-responders?	Three-way interaction between the two initial factors and the subsequent factor for non-responders, averaging over response status	2θ7