Although one of the most important elements of most projects, supervision is also one of the weakest. There are several reasons for this. First, supervision is not given an adequate place in program plans and designs. A solid study of successful programs suggests that a key characteristic of these programs is a strong supervisory component, with a ratio of supervisors to units under supervision of between 1 to 10 and 1 to 15, depending on the circumstances (Heaver 1988). The study recommends a ratio of 1 to 12. In theory, this would allow a supervisor to visit each unit at least twice a month, even if a visit requires a full day. This ratio may be too low for a population concentrated in an urban area, where a supervisor may be able to handle more units. But it may be too high for a dispersed rural area.

Second, supervisors do not get to the field often enough, a problem that may have little to do with the number of units a supervisor is responsible for. Infrequent visits may instead be related to transportation problems. Project vehicles may be lacking or broken down, or there may be a lack of gasoline. Or vehicles that are supposed to be at the disposal of supervisors may be unavailable because they are being used by higher-level administrative staff. And sometimes relying on public transportation proves impossible, or the project fails to budget adequately for transportation expenses, which then must come out of the supervisor's pocket.

Motivational problems may also play a part. Some supervisors remain in their offices rather than visit the field because they prefer not to travel and see their position as a sinecure. Sometimes the lack of motivation is related to salary.

Third, supervisors traditionally have been given the responsibility of "inspecting" those they are to supervise. Defining their role as one of inspection is very different from defining it as one of providing continuous training on the job or of "accompanying" those implementing programs on site. Supervisors need training to learn a more educational and supportive role.

Finally, the selection process is often faulty. Many supervisors are selected for political reasons or on the basis of criteria that have little to do with practical knowledge of child development or of how to function in a child care and development program. These supervisors then find it difficult to provide the methodological and content support needed.

Experience has identified several innovations that strengthen the supervisory role:

• Encouraging biweekly or monthly meetings between caregivers and their supervisor for joint planning. Such meetings help identify caregivers who need more help than others, give caregivers the opportunity to provide others with solutions to problems, and allow the supervisor to call on specialists to make presentations or work with practitioners in a group (such as a medical doctor who can respond to questions about health problems that a supervisor may be unable to answer).

• Recruiting supervisors from the ranks of the best practitioners, disregarding formal educational requirements that might disqualify otherwise highly qualified candidates. (This possibility for advancement also helps to motivate practitioners.)

• Varying the ratio of supervisors to units over time so that a supervisor has fewer units during the startup and more as the field units gain experience. Initially the ratio might be 1 to 10 or fewer; later it could rise to 1 to 20 or more. Or a project could hire a mix of new and more experienced practitioners, so that more experienced people could help the supervisor during the regular meetings.