1. It is not a substitute for balanced reading instruction

Whatever its advantages, AR should clearly not form the basis or core of the reading curriculum. It is supplementary and complementary to balanced reading instruction, including instruction in phonics, and not a substitute for it. It is just another tool to help the teacher deliver the curriculum effectively, albeit a powerful one under the right circumstances.

2. It is not a “reading incentive” program

Some teachers choose to use the promise of extrinsic reward as an incentive for some or all of their students to earn points on the AR tests. Assertions in the media hold that extrinsic reward harms students' intrinsic motivation, but there is no evidence to support these claims (Cameron & Pierce, 1994). There are, however, large cultural variations in the acceptability (and associated effectiveness) of this practice. Certainly extrinsic reward for reading would generally be considered inappropriate, if not unacceptable, in Europe. Indeed, a study by Vollands, Topping, and Evans (1999) found that even socioeconomically disadvantaged elementary students in Scotland were totally disinterested in any tangible rewards, but they were highly motivated by the individualized performance feedback inherent in the program. In my experience, cultural differences also exist within the United States, with extrinsic rewards seeming to be more widely used by teachers in the South. Likewise, there are differences in the extent to which teachers post students' progress in point acquisition in a public area of the classroom for others to see.

The Accelerated Reader does seem to motivate many students to read more, and to that extent it could be considered at least partially a “reading motivation” program. However, there is nothing whatsoever in the software that requires the attachment of extrinsic rewards for reading. The decision of whether to use a reward system is a matter to be determined purely by the professional judgment of the teacher, who knows his or her students best. In schools where rewards are attached to AR points, many teachers wisely choose to make those rewards books or vouchers for books. For general discussion of the usefulness of books as rewards, see, for example, McLaughlin (1975).

3. It is not The Electronic Bookshelf

The Electronic Bookshelf (1997) program has many similarities to AR -- but also some very significant differences. Everhart (1998, 1999) provides a useful comparative analysis of the operating features of AR, EBS, and other related software, but such information inevitably dates fairly quickly.

Test retaking. The Electronic Bookshelf allows test retaking; the AR default condition is that students can test on a book only once, since failure on a test probably indicates that the book has not been read or was too difficult. (The AR “Literacy Skills” tests, described under What Is The Accelerated Reader?, are an exception. These tests are not intended primarily to motivate students, are based on a large item bank, and can be used for pre- and post-testing.)

Point value. In order to maintain fairness and consistency, AR does not allow the teacher the option of assigning different point values for specific books. The Electronic Bookshelf does allow point values to be changed by the teacher.

Pass criteria. Again in order to maintain fairness and consistency, AR, unlike EBS, does not allow the teacher to change the number of correct answers required to pass a test. Further, EBS only records pass or fail, rather than graduated performance. Research has demonstrated that in AR, deep comprehension of a book resulting in value added to students' tested reading performance is closely associated with a minimum of 85 percent correct responses on the accompanying AR quiz (Sanders & Topping, 1999; see also Research on the Accelerated Reader). Therefore, EBS users deploy lower pass rates at their peril.

Some teachers prefer EBS to AR precisely because it seems to give them flexibility to meet the needs of individual students. Of course, when that flexibility is exercised, the psychometric properties of the system are severely damaged: much of the information collected is rendered meaningless because data are not comparable from student to student, and it becomes very difficult to manage the system consistently and fairly (see Psychometric Properties).

4. It is not Reading Counts

Early in 1999, the relatively small EBS operation was taken over by Scholastic, which is marketing a refurbished version of the program under the name Reading Counts (RC) and connecting software sales to sales of other Scholastic products.

According to the information placed in the RC area of the Scholastic website at various times over the past months, the RC software features an “exciting graphic interface”: though test presentation itself seems relatively plain, the “You have passed” screens are bright, cheerful, and varied. In addition, after passing a test, students are led to a screen that asks them to rate the book on a simple 5-point scale. Over time, the system reports on class favorites. Whether seeing “their names up in lights” enhances or impairs students' self-monitoring of improvement in reading comprehension might, of course, vary from student to student.

Reading Counts offers an “extensive database of over 15,000 books and quizzes.” Readability levels are assigned for each title based on the Lexile framework (AR uses Flesch-Kincaid, but there is little evidence that one is better than the other). If students do not pass a test, an advisory screen “prompts them to review the book and try again.” Up to nine successive retestings are thus actively encouraged.

The RC program incorporates a modest degree of item banking, with combinations of (“in general”) 30 items available for each test. As it is virtually impossible to create item banks with stable psychometric properties for very short and easy books, teachers of the early grades will wish to exercise caution. In the upper grades, a 10-item test drawn from a 30-item bank clearly offers few alternative permutations, so teachers will again wish to exercise caution. In any event, successive retestings on an individual title will swiftly exhaust the 30-item bank.

For busy teachers, RC's “Auto Alerts” are potentially a most useful feature. Unlike AR, in which the teacher must request generation of an at-risk report, RC automatically generates on-screen messages when the teacher logs in, offering a reminder to intervene with students who were flagged as at risk in the previous session.

Scholastic describes RC as research based, and indeed the company makes a Research and Results Report available on request. However, the report proves to consist of a summary of general findings about reading achievement (e.g., NAEP data) and brief, rather anecdotal accounts of uncertain vintage on the use of EBS, which fail to demonstrate any causal connection between the software and improved reading comprehension (as Scholastic puts it, “the following outcomes in case studies are attributed [emphasis added] to the Electronic Bookshelf”). No references to the full text of the case studies are supplied, and no specific research studies of either EBS or RC are offered. Indeed, my own search of the literature failed to locate any significant published evaluation of EBS, although Carver and Leibert (1995) used EBS in a study of the impact of reading challenge, and reported on difficulties such as low validity in the EBS book difficulty leveling and students testing before they had finished reading the book.

5. It is not Book Adventure

Another recent development is Book Adventure, described at its website as a “free, Web-based reading incentive program dedicated to helping kids read, click, and win!” As of this writing, Book Adventure offered online tests for 3000 books, and 7000 more are planned. Registered participants can access the site from home or school and use the search engine to find titles of interest to them for which Book Adventure tests are available. After finding and reading a selected book, the student then returns to the site to take the quiz, earning points that can be traded in for prizes. The website states that “as soon as a quiz has been completed, the Reader sees his or her points displayed based on how many quiz questions were correctly answered. Each quiz can be taken twice; only the highest score is recorded. The Reader may then re-take the quiz, take a quiz on a new book, or, best of all, check out Prizes! Some of the prizes are donated by Book Adventure Sponsors, while others are prizes that can be printed off immediately!” The percentage of correct answers required to pass a test and accumulate points appears to be very low, at least for some tests. It is not stated whether the prizes are available outside the United States.

Book Adventure is supported by corporate sponsors, including the book publisher Houghton Mifflin and the bookstore chain Barnes & Noble. A number of “Educational Partners,” including the International Reading Association (publisher of this e-journal), are named at the site, although their role is unspecified.

Book Adventure claims only to be a “reading incentive” program and, for better or worse, offers extrinsic rewards for volume of reading. Unlike AR, it can provide no stable, reliable, or valid information to evaluate students' progress in reading achievement or to serve as part of any formative assessment. It does, however, present an inexpensive option, albeit one restricted to a very small number of books at each grade level at the moment.

Go to map