Advantages and Disadvantages

To avoid the possibility of prejudices running amok, assumed or theoretical advantages should be carefully distinguished from those actually experienced in practice or evidenced by research. Following are advantages and disadvantages -- real and perceived -- of the AR program. (For an earlier analysis of the AR software's strengths and weaknesses, see Poock, 1998. Although the article is based only upon her own experience as a librarian in two elementary schools, it is nonetheless interesting reading. She notes “there are many misuses of this program.”)

Advantages Disadvantages
1. Students test independently
Students soon learn to take AR tests independently, unless they are very young (first graders) or special education students.
1. Exceptional students
Using AR with very young students, delayed readers, or those with special needs (e.g., students with dyslexia or attention deficit disorders) requires more effort and organization. Support is needed for reading both the book and the test, without the helper being drawn into giving the testee the correct answers. (Tests do give helpers a scaffolded framework that can lead to discussion of the book.) Helpers can be teaching assistants, adult volunteers, and same-age or cross-age peer tutors, but time is needed for training and scheduling.
   Delayed readers might not want to be seen reading low readability books, and they may therefore choose books that are too difficult for them. This problem needs handling at the source, not by downward adjustment of the pass criterion. Gifted students might not enjoy the program at all, especially if they are already highly motivated and extremely competent readers (although Peak and Dewalt, 1993, reported good results with able readers). This raises the issue of whether involvement in AR should be voluntary or compulsory, for some or all students.
2. Number of books covered
As of this writing, AR tests are available for over 25,000 books, including fiction and nonfiction, and English- and Spanish-language titles. Books are recommended by teachers and librarians, and the software publisher welcomes recommendations of books for which no AR test exists. Test libraries customized to a school's book stocks are supplied. Teachers also have a shell facility that enables them to write and integrate tests of their own making on books for which no test exists. However, this takes teacher time, and the psychometric properties of teacher-made tests are unlikely to be as stable as those created by the software publishers.
2. Student preoccupation with AR books
Some teachers are concerned that students might not read more, but only more narrowly, focusing on books for which tests are available. This is a particular concern when AR is first purchased and has high novelty value, which is often also the time when only an initial small batch of AR tests has been acquired. A strategic decision is required about whether AR books should be indicated as such (e.g., with a spine sticker); certainly, keeping AR books separate from other titles in the library is not appropriate. Some school libraries impose a rule that a child can check out a pair of books -- one AR and one not. After the school has audited its book stocks and ordered a larger customized AR test collection to fit, such concerns usually dissipate.
3. Increase in library circulation
There are anecdotal accounts of school library circulation figures increasing as much as fivefold after the introduction of AR (see, e.g., Poock, 1998).
3. Exhaustion of library resources
Greatly increased circulation can stress the library's book supply and check-out system. Teachers soon start feeling that they need more books, and links to outside library services might have to be explored. The program might also generate a feeling that books should be coded for readability (though many schools do this anyway).
4. Empowers teachers to intervene effectively
With AR, teachers can identify at-risk children and help them get back on track quickly.
4. Teacher stress
Having better quality information about what is going on in class means that the many possible interventions that need to be made with students are more readily apparent. This pressure might be what leads some teachers to alter the structural “rules” of AR for particular students, thereby invalidating its psychometric properties.
5. Can increase motivation and achievement
With good implementation, research strongly suggests that such increases are seen (see Research on The Accelerated Reader).
5. Only good implementation increases achievement
There are many programs in use in education that have not been shown to be effective, however they are implemented; for many others, it is unclear what good implementation might look like. With AR, good or bad implementation is not a dichotomy, but a matter of degree. As the Vollands, Topping, and Evans (1999) study showed, less than perfect implementation of AR can still yield gains.
6. Focus on literal comprehension
As outlined under Psychometric Properties, AR test questions focus mainly on literal comprehension. These questions yield adequate reliability and validity for the purpose of measuring reading practice.
6. Focus on literal comprehension
For the assessment of idiosyncratic reader inference in a more open-ended way, teachers must generate their own questions, ask students who have read a particular book to generate questions for one another, or, if they own version 5.0 of the software, use the “Literacy Skills” tests available as an extension.
7. Technical support
Detailed technical support is available from the software publishers through a toll-free (in the U.S. and Canada) telephone number. Users can also subscribe to a listserv where they can exchange information with teachers at other AR-owning schools.
7. Technical concerns
Access to hardware can be problematic. In schools with computer laboratories rather than classroom-based hardware, access is often time-tabled and routinized, and AR can therefore not be integrated with the reading experience in the classroom. Installation of the software and familiarization with it takes time.
8. Cost-effectiveness
The highest level of cost-effectiveness comes in schools where the program has been in use for some time, operated in a coordinated way in many classes (especially through a school network), and where a large number of tests closely linked to book stocks has been acquired. Very small rural schools might struggle to reach such a critical mass.
8. Cost
Like most software, AR costs money. Of course, most things in education cost money, and given limited resources, difficult decisions about purchasing new materials must be made.
9. Teacher training
Training is available at different levels of intensity. The Reading Renaissance training associated with AR aims to improve implementation quality.
9. Cost in terms of teacher time
Like any other educational program, AR has to be managed effectively by professional teachers -- and thus it incurs time costs. Time is needed for installation of the software, training, auditing and managing book supply, scrutinizing AR data and reports, and intervening with students. It is realistic to expect that good implementation of the program might not save teacher time, but it should enable more effective use of that time.
10. Quality assurance
The Model Classroom program associated with AR offers certification for meeting precisely specified minimum implementation standards.
10. Instructional substitution
Although AR is intended to be supplementary and complementary to balanced reading instruction, one encounters anecdotes -- impossible to verify -- of schools where, over time, reading instruction is curtailed, perhaps on the assumption that AR and school librarians together will suffice.
11. Parental involvement
The program generates reports for parents, who are encouraged to help with reading at home. A “Family Reading” kit, which outlines other ways of linking home and school reading activities through AR, is also available.
11. Cost in terms of other curricular areas
For good implementation, extra class time must be allocated to reading. This implies that time must be taken from some other area of the curriculum (it should, however, be pointed out that nonfiction AR books can be used across the curriculum, as indeed can some of the fiction).
12. Link to peer tutoring
The program facilitates and links with same-age or cross-age peer tutoring in reading. This is further explicated in the training and supplementary publications.
12. Competition and student stress
There is anecdotal evidence that, in some schools (perhaps particularly in those where competition and extrinsic reinforcement are emphasized), some students become preoccupied with gaining AR points or with the superficialities of the AR system, rather than with reading interesting books. Too much emphasis on goal-setting can make the whole thing a numbers race. Some students read and take tests on many easy books in order to accumulate points (this is, of course, flagged by the system and a matter for teacher intervention). Some students have difficulty coping with not having the most points or always being at the bottom. All of this is likely to take the joy out of reading.
While there might be one or two such students in every class, the teacher has a profound influence on the overall ethos and needs to structure the operation of AR from the outset to take students' needs and peculiarities into account. Starting simple and building up gradually is likely to be safer than a full-tilt total bells-and-whistles onslaught from the outset. If student progress with AR is posted publicly, it might include only names of students who have shown the largest percentage gain in the week over their own previous performance.
   Some teachers ask about how to use AR testing results as part of a student's grade. In my opinion, AR is unlikely to raise reading motivation broadly for the long-term if it is linked to high-stakes issues such as grades. Indeed, doing so seems likely to damage the intrinsic motivation that stems from feedback provided directly to the student. Using a computer-assisted norm-referenced test of reading competence makes much more sense for grade determination. But then, we are all right about things that are a matter of opinion. Sounds like an issue for debate in the Online Discussion Forum.

Go to map

Reading Online,
Posted November 1999
© 1999-2000 International Reading Association, Inc. ISSN 1096-1232