Modeling the convection of stellar interiors is crucial for stellar evolutionary studies. During the last decades different approximations (phenomenological, simulations, etc.) have been developed in order to improve the convection description. In spite of that, most of the stellar evolutionary tracks or isochrones are built, basically, using a theory established more than 50 years ago: the famous mixing length theory (MLT). It is very well known the MLT limitations and its inconsistencies. However MLT is still strongly used. Probably for good reasons (e.g. good predictions for the solar model and user friendly implementation in stellar evolutionary codes). The computation of stellar evolutionary models using MLT is dependent, at least, of two free parameters: the MLT parameter (particularly important for the super-adiabatic layer, in convective envelopes) and the amount of overshooting in convective cores. In this framework, the question is: are MLT parameter or overshoting constants from star to star (independently of the stellar mass, chemical composition and evolutionary stage)? To help to answer this question the FGK stars members of binaries play an important role. In this presentation, we point out the main contributions of binaries for the convection parameters estimation, in particularly for the mixing length parameter.