Interpretation Rules
How to Read This Data Honestly
Data is only useful if you read it honestly. Before drawing conclusions from this survey, here are the rules that should guide your interpretation — and a clear statement about what this research is and isn’t. These rules aren’t caveats. They’re the conditions under which the data is meaningful.
Five rules govern how this data should be read, quoted, and applied. They protect the integrity of the research and ensure that findings are used to strengthen governance — not distort it. If you’re citing this research in a report, briefing, or public communication, start here.
The Five Rules
Five Rules for Honest Interpretation
1
This is a snapshot, not a trend
This survey captures a single point in time — fieldwork conducted between 14 November 2025 and 25 January 2026. It tells you where these 136 organisations are now, not where they were last year or where they’ll be next year. Use it as a diagnostic mirror, not a league table.
The survey is intentionally positioned as the first in a potential series, enabling longitudinal comparison of organisational responses over time. But this edition stands alone. Read it as an indicative organisational insight base — not a population estimate.
2
Self-selecting means self-aware
Respondents chose to complete this survey. Participation was voluntary and open, distributed via professional networks, sector bodies, equality and human rights organisations, and direct outreach. No weighting was applied.
This means the sample almost certainly over-represents organisations that are already thinking about trans and nonbinary inclusion — those with EDI infrastructure, engagement with sector networks, or at least one person who cared enough to complete a 39-question survey. The reality across UK organisations as a whole is likely worse than these numbers suggest, not better.
The resulting dataset is a self-selecting, non-probability sample. It should be interpreted as an indicative national insight snapshot, not a statistically representative estimate of all UK organisations.
3
This is organisational self-report, not employee sentiment
The survey captures what organisations report about themselves — from respondents speaking from their role and experience. Several measures are perceptions and estimates, not direct measurements of employee experience.
For example, questions about how comfortable trans or nonbinary staff would feel disclosing their identity, or how confidently managers would apply policy in practice — these function as risk indicators, revealing what respondents believe their organisation would allow, tolerate, or handle. They are not definitive employee-level metrics.
Use these findings to test systems, governance, and decision pathways — not as a basis for claims about what all employees feel or do.
4
“Don’t Know” is a signal, not a gap
When a respondent says “Don’t Know”, that’s data. It tells you that the person responsible for answering — typically someone in HR, EDI, or people leadership — doesn’t know whether their organisation has a policy, a process, or a practice in a given area.
That uncertainty is itself a governance finding — and often a more revealing one than a clear “No”. A “No” means the organisation has assessed the question and concluded it doesn’t do something. A “Don’t Know” means nobody is tracking it at all.
“Don’t Know” responses are treated as data throughout this research, not as missing values. They are often signalling visibility, measurement, or governance gaps rather than neutrality.
5
Not all questions share the same base
The survey used conditional display logic, meaning not all respondents were asked all questions. Policy follow-ups were hidden where no policy existed; pushback follow-ups were shown only if pushback was reported.
This means percentages are based on the applicable base for that question (not always n=136). Where questions are multi-select, totals may exceed 100%. And “blank” is not automatically “no” — it may reflect skip logic, non-response, or partial completion.
Where a specific statistic has a different base size, the report notes this. When reading any figure, consider: who was asked this question, and what were the conditions?
Survey at a Glance
SAMPLE SIZE CONTEXT
Total responses received: n = 138
In-window responses analysed: n = 136 (two responses fell outside the stated fieldwork window and are excluded from headline reporting)
Fieldwork window: 14 November 2025 to 25 January 2026
Survey instrument: 39 questions (Q1–Q36 quantitative; Q37–Q39 qualitative open-text)
Design: Mixed-methods — structured quantitative questions with open-text qualitative prompts
This is a meaningful sample for exploratory insight research — large enough to surface repeatable patterns across governance, leadership behaviour, and operational infrastructure. But it is not statistically representative of all UK organisations. Treat these findings as directional signals that inform governance conversations — not as definitive benchmarks.
A “Don’t Know” from the person responsible for inclusion policy is not a knowledge gap. It’s a governance gap.
— Beyond Compliance research, 2025
Boundaries
What This Research Is Not
To support accurate interpretation, the findings must not be presented or cited in the following ways:
Not compliance evidence
This report cannot be cited as proof that an individual organisation is compliant, “safe”, “inclusive”, “best practice”, or legally defensible. It is not a certification, audit, or assurance product.
Not UK-wide prevalence
The dataset is a self-selecting, non-probability sample and must not be presented as “UK organisations overall”, “most UK employers”, or “national prevalence rates”.
Not proof of what employees feel
Perceptual/estimate items (e.g., disclosure comfort, perceived leadership posture) must not be presented as direct employee sentiment measurement or as evidence of “how trans/nonbinary staff in the UK feel”.
Not a ranking tool
Findings must not be used to rank, score, or name organisations, sectors, or locations as “good/bad”, “safe/unsafe”, or “leading/lagging” beyond what the report explicitly supports with base sizes and caveats.
Not a legal conclusion
Findings must not be framed as legal determinations, legal advice, or proof that a specific policy model is required or unlawful.
Not selective proof-texting
Excerpts must not be stripped of qualifiers (e.g., applicable bases, “Don’t Know” interpretation, multi-select totals) in ways that materially change meaning. Any public use should include the quoted statistic and its base (n), note that findings are indicative self-reported organisational responses from a non-probability sample, and — for estimate items — note that these are risk indicators, not workforce measurements.
Quoting the Findings Responsibly
This report may be quoted with attribution. Where excerpts are used publicly, they should be reproduced accurately and with context, including:
- The relevant base size (n) for the cited result, and the applicable base where conditional logic applies
- The fieldwork window (November 2025 – January 2026)
- The interpretation guardrail: findings are indicative, self-reported organisational responses, not audited practice
- Where the item is perceptual or estimated, it should be treated as a risk indicator, not workforce measurement
If in doubt about interpretation, appropriate framing, or responsible contextualisation — particularly in media, policy, or contested environments — contact Joanne Lockwood for a briefing.
Benchmark Your Approach
Now that you know how to read the data, benchmark your own organisation. Our free diagnostic applies the same governance framework — 50 questions across 5 domains, with results you can compare to the research findings.
Honest Data Deserves Honest Reading
This research exists to strengthen governance, not to score points. SEE Change Happen helps organisations act lawfully, humanely, and confidently — starting with an honest understanding of where they are and what needs to change.