The purpose of this study was to determine the extent to which the following factors predict computer use by teachers in classroom instruction: attitudes of teachers toward computers in the classroom, access by teachers and students to computers, training of teachers in computer use, support of teachers in their use of computers, age of the teacher, grade level in which the teacher teaches, curriculum area in which the teacher teaches, gender of the teacher, and number of years the teacher is from retirement. Computer use was measured in five ways: over-all computer use and use in drill and practice, whole class instruction, student-directed learning, and computer skills instruction.

The design of the study was both quantitative and qualitative. The population of the study was the classroom teachers of Carroll County (Virginia) Public Schools. A survey instrument was designed to measure computer use and the factors related to use. The responses from the survey were analyzed with multiple regression techniques to determine which factors were predictors of computer use by teachers in classroom instruction. The qualitative portion of the study consisted of five focus groups (5-7 teachers from grades PreK-2, 3-5, 6-7, 8-9, 10-12). The nominal group technique was used to create a prioritized list of strategies to improve teacher use of computers by focusing on the factors determined to be predictors.

Factors that predict computer use varied by grade level. Training was the most common predictor followed by attitude, support, access, and age of teacher. The prioritized lists of strategies from the focus groups included grade and curriculum specific computer training, technology "coaches" in every building, and computer labs in every building. A major implication of the study was that training must be specifically targeted to grade level and curriculum area to be effective.