Negative Case Analysis: Testing a Finding Against What Does Not Fit

What Is Negative Case Analysis?

Negative case analysis, also called deviant case analysis, is the deliberate search for accounts that run against your working explanation. Mays and Pope describe it as a long-established tactic: search for, and discuss, elements in the data that contradict, or seem to contradict, the emerging explanation, so the analysis can be refined until it explains all or the vast majority of the cases. Pope, Ziebland and Mays add that analysis alongside data collection lets the researcher look for deviant or negative cases that run counter to emerging propositions and use them to refine those propositions.

It is a check on the analyst, not on the participants. A case that does not fit is information about the explanation.

Decision This Supports

It helps decide how strongly to word a finding. If a negative case reveals a second pathway, the finding is revised. If it shows that the claim holds only under some conditions, the claim is narrowed. If it cannot be explained, the finding is stated with that exception visible. This keeps a recommendation from resting on a tidy story.

Before You Start

• Excerpt IDs and a codebook, so every case can be found again. See qualitative codebook.
• A written version of the working claim, specific enough to be wrong.
• A willingness to ask each interviewee, or each transcript, "does this account fit?" and not only "does it support?"

Steps

1. State the claim in one sentence.
2. Define what a contradicting account would look like before searching.
3. Search every record for it, including ones coded elsewhere.
4. For each hit, record the excerpt ID, what it contradicts and a first explanation.
5. Choose: revise the claim, add conditions, or report the exception.
6. Write the decision in an analytic memo.

A Small Worked Example

This example is fictional.

Working claim: "People leave the evening circles because the 6 p.m. start conflicts with work or care." Looking for contradicting accounts: someone who left though the timing was fine, or someone who stayed despite the conflict.

• Excerpt E5 (P5, left): "Honestly the timing was fine. I left because the reading level was too hard and I did not want others to see."
• Excerpt E2 (P2, stayed): "I come straight from work too and I'm usually late, but the facilitator sits with me and catches me up."

E5 shows leaving without a timing problem. E2 shows a timing problem without leaving. The claim is revised: "Timing and care conflicts were common in these interviews, and they did not account for every departure; a second route involved difficulty with the material and embarrassment. Support at the session may offset timing problems." The revised sentence is longer and weaker, and it is more accurate to the six interviews. It is not a statement about all participants.

Limits and Cautions

• A search for contradictions is only as good as the sample. If no one who left quietly was interviewed, the cases may not exist in the data.
• One negative case does not always overturn a pattern. Judge it by what it shows, not by the count of exceptions. Do not turn a case into a percentage.
• Record cases that remain unexplained rather than explaining them away.
• Software suggestions tend to favor common patterns. Ask a tool to find contradicting passages, then read them yourself; see checking AI-assisted qualitative coding.
• With mixed methods, contradictions can also arise between data types; see joint displays and conflicting results.

Frequently Asked Questions

How Many Negative Cases Are Enough?

There is no number. The aim is to search honestly and show what was found, including when nothing contradicts the claim.

Is This the Same as Triangulation?

No. Triangulation compares sources or methods. Negative case analysis tests a claim within the data you have. They support each other.