How does FAN measure oral reading fluency-and how does it ensure fairness?
FAN uses new speech-recognition software to assess and record these measures of accuracy and fluency. This software recognizes accepted pronunciations of each word, taking into account, for example, variant dialects. Advantages of the speech-recognition software are that it automatically identifies many different features of speech; extracts much more information from speech than a human listener can; uses consistent procedures, time after time, to determine probabilities and produce scores; and, since it evaluates only what it has been programmed to evaluate, avoids biases based on irrelevant factors. Extensive development work was done to ensure that correctness would be measured reliably and that speakers of nonstandard varieties of English would not be unfairly penalized. For a brief discussion of development work for the new software, see the New Features section of the NAAL website. What are the differences between FAN and another new feature of NAAL, the Adult Literacy Supplemen
