The steps in the design phase are:

1. Review and make a preliminary analysis of existing information.
2. Collect information about the programme.
3. Collect data relating to the achievement of stated objectives and project performance.
4. Draw the sample for assessment of effectiveness.
5. Design the data collection (survey) instruments.
6. Field test the survey instruments.

1. Review and make a preliminary analysis of existing information

It is important to assemble a great deal of information prior to the field work phase of an evaluation, A first step is the gathering of general information of the country and the health care system (see section 1 in "PROCEDURES").

A second step should be collection of macroeconomic and social data influencing the programme (see section 3 in "PROCEDURES").

2. Collect information about the programme

All the information about the programme should be collected before the beginning of the field work phase. This might include:

a) Policy statements, circulars and ordinances regarding the programme, existing plans for implementation, etc. (see section 2 in "PROCEDURES").

b) Programme objective and immediate objectives, inputs, activities, outputs, indicators and assumptions of each component of the programme (see section 4 in "PROCEDURES").

c) Any past evaluations of the programme or its separate components

d) All relevant monitoring data.

3. Collect data relating to the achievement of objectives and project performance

Considerable time will be saved by the evaluation team if data related both to the achievement of objectives (as reflected by selected indicators) and project performance (as reflected by inputs achieved, activities undertaken, and outputs produced) are collected beforehand. Objectives, assumptions, indicators, etc. should not be altered at this stage, either in order to reflect the current status or in anticipation of the evaluation findings.

4. Draw the sample for assessment of effectiveness

The selection of a sampling scheme requires that a balance be struck between cost and precision and that the scheme is feasible given logistical and time constraints. The proposed assessment of effectiveness (see section 5 in "PROCEDURES") is not very detailed and can be carried out in a short period (7 days) at the health centre/dispensary level. Should it be decided at a later stage to conduct an evaluation at all levels of the health care system, e.g. community level (village health workers, community leaders, home level), health centers/dispensaries and district hospitals, the question of drawing a sample becomes much more complicated.

Note, the selection of the sampling scheme is a prerequisite for determining the manpower needed for the field work phase. During that phase, enough staff must be on hand to complete the field work within the designated time. Estimates of staff required can be made only when the number of sites to be visited and their relative accessibility are known. Also, transportation requirements and budgetary needs can be fixed only after the sampling scheme is defined and the staff needs determined.

5. Design the survey instruments

Generally, this step "pins down" the exact nature of the evaluation by defining exactly the field data to be gathered. Within the context of a programme evaluation, it is necessary to balance the breadth of each instrument against the depth, the desire for quantitative data against the need for quick and easy analysis, and the flexibility afforded by open-ended and opinion type questions against the clarity and simplicity of highly structured coded questions.

6. Field test the survey instruments

If a survey is to use newly developed untried instruments of data collection, then they should be field tested first. Such a test should:

a) confirm the feasibility of the package of survey instruments, especially the clarity of each question and the timing required to complete the entire process of data gathering;

b) confirm the feasibility of using the entire package of survey instruments in a selected geographical area given the timing and other logistical constraints.

Clearly, the field test could lead to appropriate modification of the original survey instruments.