A study of programed instruction sought to establish an attribute by treatment interaction (ATI) between prior familiarity of material and response mode. Two experimental variables (familiarization and response mode) and two subject attributes (sex and I.Q.) were employed. Junior High (JH) and graduate student (GS) were assigned to familiarization (F) or non-familiarization (NF) groups, and to constructed response (CR) or reading (R) modes. The F group received advance familiarizing materials. Main effects for the college data were insignificant, but sex and I.Q. were significant for JH Students. There was ATI between familiarity and response mode. F. led to increased achievement for the R group, but to decreased achievement for the CR group. The familiarity-by-response mode interaction was complicated by interacting with sex for GS and I.Q. for JH. Delayed posttest interactions were more significant than immediate results. The negative effect of F on CR achievement was felt due to lower motivation for able students to attend to familiar material. Sex results were explained by the hypothesis that passive females attend to all tasks, impulsive males only to interesting ones; thus F led to reduced male achievement. Increased retention over time was interpreted as reflecting that retention is a function of initial meaningful learning. (Author)