Purpose An important predictor of postsecondary academic success is an individual's reading comprehension skills. Postsecondary readers apply a wide range of behavioral strategies to process text for learning purposes. Currently, no tools exist to detect a reader's use of strategies. The primary aim of this study was to develop Read, Understand, Learn, & Excel, an automated tool designed to detect reading strategy use and explore its accuracy in detecting strategies when students read digital, expository text. Method An iterative design was used to develop the computer algorithm for detecting 9 reading strategies. Twelve undergraduate students read 2 expository texts that were equated for length and complexity. A human observer documented the strategies employed by each reader, whereas the computer used digital sequences to detect the same strategies. Data were then coded and analyzed to determine agreement between the 2 sources of strategy detection (i.e., the computer and the observer). Results Agreement between the computer- and human-coded strategies was 75% or higher for 6 out of the 9 strategies. Only 3 out of the 9 strategies-previewing content, evaluating amount of remaining text, and periodic review and/or iterative summarizing-had less than 60% agreement. Conclusion Read, Understand, Learn, & Excel provides proof of concept that a reader's approach to engaging with academic text can be objectively and automatically captured. Clinical implications and suggestions to improve the sensitivity of the code are discussed. Supplemental Material https://doi.org/10.23641/asha.8204786....

Read, Understand, Learn, & Excel: Development and Testing of an Automated Reading Strategy Detection Algorithm for Postsecondary Students

10.1044/2019_AJSLP-18-0181