Social programs have to deal increasingly with their effects and are required to prove them. The aspect of long-term effects is often neglected. The present study investigated the ques-tion of how short- and long-term outcomes can be measured by an impact assessment on the mentoring program ‘Nightingale, focusing on two of its main objectives: the increase of the educational motivation and the strengthening of self-esteem. A research design has been developed, which is able to show short and long-term changes (outcomes) using a quasi-experiment design and the double-difference method. By involving a comparison group, it allows the outcomes to be causally linked to the program. This non-memory-based approach takes into account both intended and non-intended personality-oriented out-comes. With the ‘five-factor model (Bleidorn and Ostendorf 2009) and the ‘self-value scale (Schütz and Selling 2006), established instruments were chosen that correspond to the scientific main quality criteria. This research design has then been discussed with rele-vant stakeholders on acceptance and practicability, based on qualitative, partially structured expert interviews. Information resulting from the interviews were extended with selected data deriving from analysis of current evaluations of ‘Nightingale. Thereby both, the ac-ceptance and the feasibility of the research design, could (conditionally) be confirmed. The work presents the possibilities and limitations regarding planning and hypothetical imple-mentation of a long-term impact analysis in the field of social economy.