The control group sample contained the same organizational levels but with employees who
joined the company but who did not take co-op. We checked the possibility of comparison
between the two samples with a high level of description (gender, study level, duration of
employment). The samples have been selected with a census method and following the
advice of Henry (1990) for dealing with small samples, for the longitudinal sample we
selected all co-op students employed at the company for the sample