“…Out of the 269 freshmen who initially volunteered a total of 83 pairs (N = 166) were created with 103 freshmen finally discarded due to the lack of an appropriate match. Pairs were sorted and matched according to two set of control variables (Ato, López, and Benavente 2013), usually associated with academic success or failure (Barbera et al 2017;Chen 2012;Fong et al 2017;Kuh et al 2006;Laskey and Hetzel 2011;Pascarella and Terenzini 2005;Richardson, Abraham, and Bond 2012;Tinto 1993Tinto , 2012Vossensteyn et al 2015): (a) academic: degree program, program year, group, subjects and number of credits registered, type of upper secondary education track, number of times that has taken the university entrance exam, place of choice of the degree program in the pre-enrolment, GPA obtained in upper secondary school, GPA obtained in university entrance exam, GPA of access to the university, dropout and change of previous studies, number of previous retakes, level of social skills, level of learning strategies and motivation, and level of commitment; and (b) demographic: age, sex, marital status, nationality, employment status, people sharing the household, socioeconomic status, and source of studies funding. The minimum required total sample size for PTP 3.0 was previously calculated (N = 102), as we report later on the statistical analysis session, which forced us to discard 32 pairs due to the limited number of student-tutors available (N = 50).…”