Skip to main content

Table 2 Lessons learned during selection process development

From: Case study of a method of development of a selection process for community health workers in sub-Saharan Africa

Stage

Process

Lesson

1

Review existing selection process

An existing process (or processes) provide(s) ideas and examples but may not necessarily be effective even within the setting for which they were designed: evaluating predictive validity is important.

2

Observational job analysis

Use each household visit/client contact/CHW activity as the “unit” if such activities are all of a similar duration, or a suitable time block (e.g. 10–15 min) if the activities undertaken vary in duration.

Include data collection on the type of activity as this may help the choice of scenario topics for questions developed.

3

Obtaining views of stakeholders

Include CHWs if pre-implementation testing is limited—we obtained a considerable amount of feedback from CHWs during the process of training field workers to undertake the cognitive interviews.

3

Data analysis of sorting task

A different approach (to that planned and therefore used here) may be warranted given that no KSAs were identified for inclusion based on the data from Ghana. For example, excellent and important ratings could be combined, or stakeholders could be asked to select the five most important KSAs.

3

Managing stakeholders’ expectations

Stakeholders may have unrealistic expectations of the resources available for selection in terms of how many people should sit on an interview panel (a panel with multiple interviewers may also be unacceptable to CHWs, with some reporting in alpha and beta testing that any kind of interview would be very stressful).

4

Writing high quality questions

Allow sufficient time and be prepared for the process to be a little challenging.

4

Determining and operationalising marking/scoring criteria for interview questions

Criteria need to be explicit and interviewer training is essential to ensure fairness across CHWs interviewed by different individuals.

5

Determining test length

The length of test that is feasible for CHW programmes may not be sufficiently reliable and thus a trade-off between reliability and feasibility may be required.

5

Deciding how to combine scores from different elements of a selection process

Given the low correlations between scores on each element, a low passing score for the written test may be appropriate, followed by combining scores for final selection decisions. However, this may increase the number of applicants shortlisted for interview to an unmanageable number.

5

Ensuring content validity

Keep questions as simple as possible and double-check that they relate to CHWs’ on-the-job roles (with CHWs if possible).

5

Enhancing fairness and applicant acceptability

Provide applicants with information about the selection process in advance, possibly including some example questions.