Experimental analyses for improving reading fluency deficits have rarely targeted generalized increases in academic responding. As a consequence, the variables that may help students to generalize newly learned forms of academic responding like reading are not well understood. Furthermore, experimental analyses of reading fluency interventions have not systematically examined difficulty level as a variable that may affect instructional outcomes. The experiment reported in this paper expands (a) the measurement of the dependent variables to include generalized increases across tasks (reading passages) and (b) the combination of independent variables used to produce measurable generalized increases. The results demonstrate the importance of combining reward and instructional variables (including difficulty level) to produce generalized increases and how those variables can be meaningfully investigated prior to making treatment recommendations.

The technology of brief experimental analysis is just beginning to be used for identification of effective treatments for individual students who experience difficulty with oral reading fluency. In this study, the effect of a reading fluency treatment package was examined on easy and hard passages, and generalization was assessed on passages with high content overlap. The results suggest that the treatment package increased reading fluency for all 3 students. Effects were moderated by difficulty level for all 3 students. Results are discussed in terms of future refinements to the procedures, validation of the methods, and potential applications in clinical and school settings.