Key Informant Interview

Some things about a project are known only to a handful of people. The doctor who ran the camps knows why two of them were cancelled. The panchayat member knows which hamlet was left out and why.

A Key Informant Interview, or KII, is how an assessor gets at that knowledge. It is a one-to-one conversation with a person who holds specialist knowledge of the project.

The workbook puts interviews among an assessor's core data collection tools. They can be structured, following set questions, or semi-structured, leaving room to follow up and probe. They suit in-depth and personalised information. And the workbook says they are often used to collect data from key informants, experts, or individuals with unique perspectives.

A KII is not a survey and not a group. A survey gives breadth — hundreds of answers to the same question. A focus group discussion gives a community's shared view. A KII gives one informed person's account, in depth.

Every case study in Chapter 8 uses all three together. The counts are small: 8 or 10 KIIs against 20 focus groups and 500 to 750 surveys.

Who counts as a key informant. Across the Chapter 8 case studies the informants are people with institutional or professional knowledge, not ordinary beneficiaries:

The three tools together. Every case study names the same combination — Key Informant Interviews (KII), Focus Group Discussions (FGD) and quantitative surveys — and the scale of each is deliberately different:

Small numbers of KIIs are not a weakness. The point of a key informant is that few people have the knowledge, so interviewing more would add little.

Where KIIs sit in the assessor's sequence. The case studies follow a consistent order: identify stakeholders, then desk review of existing documents, then physical inspection, then personal interviews and group discussions. The KII therefore comes after the assessor already knows what the documents claim — which is what makes it useful, because she can ask the doctor about the gap between the claim and the record.

The skills it demands (Chapter 5, 5.3.4). The workbook attaches four competencies to interviewing directly: effective communication with managers, staff, beneficiaries and government officials, in the appropriate language or dialect; empathy, to establish trust and create a safe space for personal stories; active listening, understanding the emotions, tone and context behind the words so that follow-up questions can dig deeper; and cultural sensitivity, so the interview is conducted respectfully. The code of conduct's non-judgmental attitude and objectivity principles apply throughout — base findings on evidence, not on the interviewer's preconceptions.

Evidence status. A KII is assessment evidence, and the usual tests apply: competence, relevance and sufficiency. It is not self-verifying. Where a key informant's account cannot be corroborated and no further evidence can be obtained, a scope limitation arises.

The figures below are the workbook's own, from the Chapter 8 cataract case study, with one illustrative finding added.

ABC Foundation's eye healthcare project ran screening camps and free cataract surgeries in rural India, reporting over 10,000 surgeries and 200 villages covered.

The assessor's fieldwork:

What the 10 KIIs add that the 500 surveys cannot. The case reports 90% attendance at follow-up appointments and an 85% improvement in vision outcomes. Surveys confirm both. But only the surgeons can explain the infrastructure gaps the report names — limited facilities and equipment in remote areas — and only the panchayat members can explain the awareness barriers, where communities did not know the surgery was available or safe.

Suppose one surgeon says in her interview that 3 of the season's camps were held without a functioning autoclave on site. No survey question would surface that. No focus group of patients would know it. It is exactly the kind of fact a key informant holds alone.

What the assessor then does. She does not report it on one person's word. She checks the equipment logs from the desk review, and inspects the two camps still running. If the logs confirm it, the finding goes into Section III as a gap. If they contradict it, she has an unresolved diverging opinion — and under the fair-presentation principle that too must be reported.

Chapter 8 (Social Impact Assessment Case Studies, 15% weightage and the paper's heaviest chapter) names Key Informant Interviews in the data-collection section of every case study, with counts of 10 and 8 against 20 FGDs and 500 or 750 surveys. Chapter 5 (10%) lists interviews among an assessor's data collection tools and attaches the communication, empathy, active-listening and cultural-sensitivity competencies to them.

Chapter 8's questions read the case studies directly. Expect: which data collection methods were used in a named case (KII, FGD and quantitative surveys), how many KIIs were conducted, and who the key informants were (doctors, NGO staff and local government representatives). The KII-versus-FGD distinction is the conceptual half of the same mark.

• A KII is one-to-one; an FGD is a group. Both are qualitative, and both appear in the same sentence of every case study, which is exactly why they get swapped.
• Key informants are people with specialist knowledge — doctors, NGO staff, officials — not a random sample of beneficiaries. Interviewing 10 beneficiaries is not a KII exercise.
• The counts are deliberately small. 10 and 8 KIIs against 500 and 750 surveys. A small number is by design, not a sampling failure.
• KIIs are not the sample. The 5% and 15% samples in the case studies are drawn from beneficiaries for the surveys; KIIs sit outside that frame.
• Interviews may be structured or semi-structured. The workbook names both, so an option restricting KIIs to a fixed questionnaire is wrong.
• KIIs come after the desk review, not before it. The case studies are consistent: documents first, then inspection, then interviews and group discussions.
• One informant's account is not a finding. It is evidence, subject to the competence, relevance and sufficiency tests, and an uncorroborated claim that cannot be evidenced leads to a scope limitation or a reported diverging opinion.