Combine director interviews with questionnaire results by treating them as complementary evidence: the questionnaire shows where the board scores, and the interviews explain why. AI links recurring interview themes to the relevant questionnaire dimensions, while human reviewers test every finding before it reaches the board.
The method works for any board, committee, or director assessment, provided the process protects confidentiality and keeps judgement with people. The sections below address the questions Company Secretaries most often raise when designing a mixed-methods board assessment.
What should happen when interviews and questionnaires disagree?
A disagreement between interviews and questionnaire results is a finding, not an error. It should be investigated and reported. Scores often reflect a director’s polite, general view, while interviews expose the specific experiences behind it. The gap usually points to an issue the board has not yet discussed openly.
Start by identifying the type of divergence. A board may rate “culture” highly on the questionnaire while several interviewees describe reluctance to challenge the Chair. That pattern suggests social desirability bias, where directors answer anonymously in writing but speak candidly when given room to qualify their views. A second pattern appears when scores are low but interviews are muted, which often signals a topic directors find uncomfortable to articulate.
Resolve the conflict through three steps:
- Check the sample. Confirm that the interviewed directors reflect the wider respondent group, including committee chairs and newer members.
- Re-read the underlying statements. Test whether the interview remarks are widely shared or isolated.
- Report the divergence openly, with a hypothesis on its cause, and invite the board to discuss it.
Boards benefit most when the evaluator is candid about such tensions. A frank account of where perception and experience differ is more useful than a smoothed average.
How does AI link interview themes to questionnaire scores?
AI links interview themes to questionnaire scores by coding interview transcripts into recurring topics, mapping those topics to the questionnaire’s dimensions, and comparing the qualitative pattern with the quantitative result. This triangulation of qualitative and quantitative data shows which scores are supported by testimony and which are not.
In practice, AI thematic analysis of board interviews follows a consistent sequence. The system reads each transcript and tags passages by subject, such as strategy oversight, information quality, boardroom dynamics, or Chair leadership. It then groups these tags into themes and measures how frequently and how strongly each theme appears across directors. Finally, it aligns each theme with the matching questionnaire items, so a low score on “quality of debate” sits beside the interview passages that describe how debate actually unfolds.
The value lies in speed and consistency. A human reviewer can hold perhaps a dozen interviews in mind. AI can surface cross-cutting patterns, including themes that no single director stated outright. It cannot judge significance, tone, or political context. Those remain human responsibilities, which is why the validation step below matters.
Which data source should lead the evaluation analysis?
Neither source should lead by default. The questionnaire should establish the baseline and identify where to look, while the interviews should explain and qualify what the numbers show. When the two conflict on a material matter, the interview evidence deserves closer scrutiny because it carries the context.
Questionnaires offer breadth, comparability across years, and a clear view of each director’s ratings. Interviews offer depth, nuance, and the forward-looking judgement that a rating scale cannot capture. A rigorous mixed-methods board assessment uses each for what it does best:
- Use questionnaire data to benchmark performance, track change over time, and spot outliers across directors or committees.
- Use interview data to understand dynamics, relationships, culture, and the reasoning behind a score.
- Use both together to test whether a conclusion holds under two different lines of evidence.
Boards that rely on scores alone tend to produce compliance-oriented reports. Boards that rely on interviews alone risk over-weighting the loudest voices. The strongest evaluations weigh both and say plainly which carries more weight in each finding.
How can directors’ confidentiality be protected during AI analysis?
Directors’ confidentiality is protected by controlling who sees raw data, removing identifying details before analysis, and setting clear rules on how the AI platform stores and uses information. Directors must be told in advance how their words will be processed, and attribution should never appear in the final report.
Key safeguards include:
- Anonymisation before analysis. Strip names, roles, and distinctive phrases from transcripts so that themes cannot be traced to individuals, particularly on small boards.
- Restricted access. Limit raw responses to the independent evaluator. The Company Secretary and Chair should receive synthesised findings only.
- Data governance terms. Confirm that the platform does not retain data beyond the engagement and does not use board content to train external models.
- Minimum group sizes. Suppress breakdowns by group where the number of respondents is too small to preserve anonymity.
- Informed consent. Explain the process to every director before interviews begin.
Candour depends on trust. If directors doubt the process, they will moderate what they say, and the evaluation will lose its value. The Company Secretary is well placed to safeguard this trust by making these protections explicit from the outset.
Who should validate the AI-generated findings before reporting?
An experienced human evaluator, ideally an independent one, should validate AI-generated findings before they are reported. This person tests each theme against the source material, applies boardroom judgement, and removes anything that is inaccurate, misleading, or identifies a director. The Chair and Company Secretary then review the findings for factual accuracy and context.
Validation should cover four checks. First, traceability: every finding must link back to specific evidence in the interviews and questionnaire data. Second, proportionality: a theme raised by one director should not be presented as a board-wide concern. Third, context: the evaluator must consider history, personalities, and pending decisions that the AI cannot see. Fourth, tone: findings should be honest and constructive, not softened to avoid discomfort.
AI accelerates analysis, but accountability stays with people. Independent validation also strengthens the board’s compliance standing with regulators and investors, since it demonstrates that the evaluation was rigorous and not merely automated. For a deeper view of how this fits within a wider programme, see our overview of board effectiveness evaluation.
How should combined results be presented to the board?
Combined results should be presented as a short narrative organised by theme, with questionnaire scores and interview insight shown side by side for each. Lead with the most consequential findings, state strengths and development areas together, and close with a prioritised action plan that looks forward rather than backward.
A clear structure helps directors absorb the material:
- Headline findings. Three to five conclusions that matter most for the board’s future effectiveness.
- Theme by theme. For each theme, show the score, the supporting interview evidence in anonymised form, and any divergence between the two.
- Strengths and development areas. Present both, since they are inseparable in a credible evaluation.
- Recommendations. Link each to an owner and a timeline, ideally within a development plan that spans two to three years.
Present the report in a facilitated discussion rather than circulating it alone. The conversation is where findings become decisions, and where a board begins to change how it works.
How The Board Practice helps with combining interviews and questionnaires in AI-assisted evaluations
The Board Practice has refined its Board Effectiveness methodology over more than 19 years, led by Dr. Victor Prozesky, who has led or contributed to board performance programmes in over 120 instances across industries and continents. Its AI-powered board platform, launching on 1 August 2026, brings that expertise to scalable, continuous evaluation. Key elements include:
- Questionnaires that boards can generate or select, then complete within the platform
- AI-powered analysis that connects scores to qualitative evidence and produces actionable recommendations
- Candid, forward-looking findings that identify both strengths and development areas
- Close partnership with the Chair to tailor the process to the board’s own context
- Ongoing tracking of board performance, supported by a multi-year development plan
To discuss how a combined interview and questionnaire evaluation could serve your board, contact The Board Practice.
Related Articles
- How do you compare the explainability of AI board evaluation platforms?
- What is stakeholder governance and how does a digital platform make it measurable?
- What is good corporate governance?
- What is the most important fiduciary duty?
- How should the board and management team work together more effectively?