This paper discusses how conclusions drawn from research with participants in adult education have been extrapolated to the whole adult population. This tradition has been present since the 1920s in Germany. Then the core question asked of adult education was, ‘do the majority of blue collar workers want a revolution and training to rule a socialist society or not’ (Hermes)? The answers given were drawn from existing participants’ preferences and critics argued that there would be substantial differences between participants and non-participants (Buchwald). Todays’ question is: ‘What do we know about those that are not functionally literate in Western societies’? Research with literacy program participants shows they are excluded from labor markets, face dissatisfying health conditions, live in segregated areas of our cities and report negative school experiences (cf. Lehmann, Fickler-Stang, Maué 2011). But large-scale studies of the adult population (Grotlüschen, Riekmann 2012) provide a different story and show that adults that are not literate and do not participate in classes, often have jobs, and remember normal school experiences. It is suggested that program participants are often structurally different from the whole population and so there is a danger in drawing conclusions by research based only on this group.