Many different online programs for math instruction/assessment exist. ALEKS stands out. If teachers or administrators misunderstand its design, their choice of other programs will be arbitrary. They will be merely victims of marketing, not arbiters of what's best for student learning.
The internal research conducted by ALEKS on its software appears to be extensive. Every topic network is constantly and automatically checked. If less than 90% of students pass any topic, then the network to the question is re-verified. In other words, after a decade of iteration, the learning progression in ALEKS is second-to-none. However, to satisfy the various state standards, additional topics are introduced into standard curricula like Algebra 1 for coverage. These additional standards add redundancy to the network at a price of increased time (more topics) to master a course.
ALEKS uses time more efficiently than any other math program. This is easy to show. First, ALEKS uses
constructed responses only, not
multiple choice. Fewer correct answers are needed to show mastery; since multiple choice inherently requires more questions to cope with
false positives due to random guessing. Second, the ALEKS web of topics and intentional time delay in verifying mastery drive efficiency. No topic is offered until the student has a 90% chance of "passing it" due to completing a network of prerequisites. Also, while usually only four correct answers are necessary to complete a topic, one more topic question is presented days or weeks later to verify mastery. A wrong answer brings the topic back. Third, its artificial intelligence engine further separates ALEKS from most other programs. It appears to use the
time-to-answer a sequence of questions correctly to determine how many questions are needed. Sometimes a student only needs to answer two questions to complete a topic (Mastery verification is later.). In short, there is no faster method than ALEKS to develop procedural math mastery.
As part of its interface design, ALEKS doesn't display information that the student doesn't need. For example, neither the time on the question nor the menus of topics is displayed; although ALEKS does record time, when the student is working on a problem. ALEKS wants students focused on the question without distraction. There is research behind this decision. Additionally, ALEKS chose to use Java extensions to provide online protractors, rulers, etc to produce its online manipulatives. This standard software can either be automatically retrieved (www.aleks.com/plugin) or installed. Most other companies use Flash for interactions. These work well, but are losing their position of being an Internet standard. In addition, ALEKS uses natural display of answers and allows for variations in the answer of questions. These features relax students and minimize their frustration - not a small issue.
The last two paragraphs covered Assessment & Interface
Design. of the six main aspects of online math software.
ALEKS receives the most criticism on its Instructional Design. In general, the criticisms are ill-considered. First, instruction is offered only when students request it. There are no hints, and students know that they will have to do more work, if they ask for instruction. This motivates students to try to solve problems first and, if unsuccessful, to
pay close attention to solutions because they don't want to learn more than once. Second, the instruction is simply text with graphics, no videos. While unappealing initially, they fit how most students actually do math. For example, few students read math textbooks. They attempt problems, then look for examples on how to do them as quickly as possible. ALEKS's instructions quickly suffice. From another perspective, video with sound appears wonderful, but it can distract and requires computers with headphones. ALEKS works quickly and quietly even in dial-up environments. More importantly, videos frequently waste time. Students just look at them in a daze. The
low density of information in a video is rarely worth the time to most students. One exception exists.
Buzzmath has annotated videos, which allows students to skip sections they don't want to see so that they can go to what they want quickly. ALEKS could offer such a service as an upgrade.
Intermixed with Assessment Design is Database Design. ALEKS satisfies the main math problem criterion: whether through randomization or question quantity, the number of questions per topic is far more than what can be shared between students. In short, cheating by having shared question/answer lists is squelched by ALEKS and several other programs. Another subtle feature is the speed of the database when reports are being generated. ALEKS is so fast, that the browser refresh rate appears slower. Other programs, such as APEX, have slow to generate reports. From a teacher usage perspective, fast reports are critical.
The bars and circle charts in ALEKS's reports are easy to understand. Many have hyperlinks to other reports which greatly saves time. The only program with better reports may be
Daskala, which is limited to simple sequences of multiple choice questions. ALEKS reporting could be improved by an all student-class report. This is a small issue, but teachers who have students spread between many classes, have to remember what class a student is in to find the student's status. It would be valuable to have a list report of students and their classes with hyperlinks to the student's report.
The one area in which ALEKS can improve, or at least freshen, is account management. ALEKS works with
active licenses, which is the best method, but ALEKS lacks ease in both enrolling students and also in identifying
unenrolled users. The information is there, but it is tedious to use. On the other hand, unlike many other active license software vendors, it doesn't lose data on previous students. A simple list with fields would resolve the enrollment issue.
The great issue with ALEKS is that it intentionally rejects rich problems - problems that require strong student insight and drawing from multiple sources to find solutions. This is contentious. Research by ALEKS probably shows what I see in the classroom: many, perhaps the majority, of independent students simply
stall out when confronted with problems that stretch them too much. This is sad, but it may just indicate that the
desire for math software to do everything - no classroom needed - is a false dream, not research-based. The 90% rule really means a great deal to ALEKS's programmers. Mastery of procedural skills is the goal. The subtleties of applications are left to classroom instruction.
A superb product with richer problems is
Carnegie Tutor, CT. Since questions are rich, each question has many objectives built into it. When student cannot solve a problem, the instructional help to get through it; including the teacher, requires considerable time, and is inherently different for each student! CT does have underlying artificial intelligence, but it is results are still linear. If a student is stuck, they must stop. This is a prescription for failure in many computer environments, but CT offers appealing instruction. Another program with rich questions is
StudyIsland. Good students are attracted to it. If a student struggles, StudyIsland, like ALEKS, allows students to work on other topics. In short, students stay working - the great benefit of student choice. StudyIsland even incorporates some simple AI to give students easier questions in a topic, if the grade level questions are causing failure. This appears wise, but the ALEKS approach of moving the student through a network of topics, not a flow of topics within the same top-level topic shows far more sophistication in improving the transferable skills of each student.
The more problems that students complete as
practice, not experience, each hour, the
better! In my observations, only SmartMath and IXL, have higher problem completion rates, with SmartMath having more concentrated practice, but these are K-6 programs! Student choice - for whatever reason, like fatigue or boredom - should be allowed. It keeps students working. Also, in years of ALEKS and SmartMath usage, not one question mistake has been found by me. APEX is riddled with errors, iPass has problems, and StudyIsland has a software link to report mistakes, to which StudyIsland responds promptly. IXL only has a few errors. This issue is missed in software evaluations. Maintaining confidence in the software is an important issue in the value of the software.
In summary, ALEKS does its jobs exceptionally well on three of six categories: Assessment, Interface, and Database. On the two categories Account Management and Reports, ALEKS is very good. Yet, strangely, in Instructional Design, ALEKS is either the best or inadequate, depending on a reviewer's knowledge and/or desires. ALEKS Corporation has decided to focus its software on what they believe works best in software, from a research basis. Others want a full, rich instructional system which is a desirable goal. This may be an example of the best being the enemy of the better. More may be desired in instructional capabilities and in rich problems, but it is not clear whether or not this is valid with more than a few students. ALEKS thinks not; and ALEKS could easily deliver additional features. In the years I've used ALEKS, the company has grown from 20 to 120 employees. More importantly, ALEKS retains its earnings and pours it into development. It is not a cash cow. ALEKS provides the fastest pathway to procedural fluency. Asking software to do more may be unwise or self-delusional.
Lastly, pricing matters. It's easy to forget that software providers are businesses that pay not only for programmers, but also for salespeople, customer service representatives, clerks, brochures, trade shows, etc. Years ago, in a private conversation with an ALEKS executive, I was told that to go below $42 average per user would kill R&D over time. Obviously, large school districts could bargain for a better price and home users, with their many customer service interactions, would have to pay more, but to maintain the average was vital. I never forget this when purchasing other software at 5 -10$ per user, which I frequently do. These programs may not be able to keep improving without venture investments, and programs that cost over $100 per year - many do - may not maintain steady users. ALEKS's pricing allows sustainability and improvement year after year.
--------------
Dennis Ashendorf has extensively used the following:
- ALEKS
- APEX
- iPass
- IXL
- RevolutionPrep
- SmartMath
- XLPrep
Dennis Ashendorf has lightly used
