One of the hottest areas for information technology in the business world is data mining: Using computers to analyze reams of consumer data and find patterns that indicate preferences, interests, and needs. Clever companies can turn that knowledge into new products and services, or change their marketing to reflect consumers’ desires and needs.

School administrators, with the help of sophisticated technology directors, are starting to use data mining to uncover “truths” about student performance in the classroom. As more student records get computerized, the opportunity to find out rapidly where students’ strengths and weaknesses lie holds great potential for providing individual teachers with information on what’s working (and isn’t working) in their classrooms.

Broward County (Fla.) makes extensive use of data mining. IBM created a system for the 125-school district several years ago, in which student-related data are entered into the system at each school and stored in a district-owned mini-computer. Now, a wide range of student information (test scores, demographics, attendance, and more) can be retrieved easily from terminals within each school.

One administrator who uses the system says the benefits are substantial. First, data can be accessed immediately, which makes it much more likely that a teacher or administrator will use the system. As a result, many more teachers and administrators are requesting information than ever before.

Second, the data can be displayed in many ways—graphs, tables, etc.—and manipulated to show cross-sections of students.

Third, although students may switch schools often, the data warehouse retains crucial background information on each student.

As an example of how data mining can change classroom teaching methods and student achievement, the author describes a situation in which a teacher found that one of his classes had extremely low scores on standardized reading tests. The teacher incorporated a slower approach (so as not to leave students behind) and more visual aids in the classroom. Tests scores rose considerably by the end of the year. However, Broward administrators note that using data mining to obtain student test scores and then basing all classroom activity on those scores would be an overreliance on the technology, since teachers know a great deal about a student’s academic achievement that doesn’t show up in test scores.