Methodology Behind the Dom/Sub/Switch Assessment

What the test is for

This assessment exists to give you language for something you may only have felt, never named. Most people arrive with something vague — "I think I like being told what to do, sometimes" — and leave with nine more specific words for parts of that feeling: dominant tendency, submissive tendency, contextual flexibility, structure, control, connection, communication, intensity, readiness. That vocabulary is the actual product. A result is a starting point for reflection and for the reading it points you toward, not a verdict on who you are.

It is built as a private mirror, not a scoreboard. There is no ranking against anyone else, no answer pattern that counts as more evolved or more authentic than another, and no attempt to tell you anything you did not already know about yourself at some level. The whole point is to make what you already sense a little easier to say out loud, even if the only person you say it to is yourself.

To answer the two questions most people actually have before trusting a page like this with anything: no, it is not scientifically validated, and no, you should not treat it as more accurate than the honest limits explained throughout the rest of this page allow it to be.

Why one slider isn't enough

The obvious way to build a test like this would be one slider, dominant on one end, submissive on the other, and a single number telling you where you land. We deliberately did not build it that way, because that model throws away real information about you. Plenty of people carry genuine comfort with both leading and following, depending on the partner, the relationship or their own state that day — a single slider forces a false either/or onto something that is often a genuine both/and.

So dominant tendency and submissive tendency are scored independently of each other, and contextual flexibility — how much your role preference actually shifts by situation — is tracked as its own dimension rather than inferred from the other two being close together. On top of that, six supporting dimensions capture how you relate to structure, negotiated control, emotional connection, explicit communication, intensity and real-life readiness, because two people can land on an identical dominant score and still want almost entirely different things from a dynamic.

The nine dimensions this test measures

Each of these is scored from its own set of statements. None of them is a personality trait outside the context of a negotiated dynamic — they describe stated comfort and interest, not a claim about how you behave anywhere else in your life.

Quick mode, Deep mode, and how you answer

You choose between two versions before you start. Quick mode uses 24 statements and takes most people four to six minutes; Deep mode uses all 72 and takes twelve to eighteen. The difference is coverage, not accuracy: Deep mode spreads each dimension across more statements, so no single answer carries as much weight and an odd or ambiguous response has less power to swing your result. That is a genuine benefit of breadth. It is not a claim that Deep mode is clinically more reliable — this tool has not gone through the reliability testing that word would require.

Every statement uses a five-point agreement scale, from strongly disagree to strongly agree, plus a separate "not sure / doesn't fit me" option that sits outside that scale rather than forcing a guess into the middle of it. That option exists on purpose. Not every statement will apply to your situation, and being unsure is a real, useful answer — it tells the assessment to leave that particular data point out of your scores rather than quietly counting an uncertain shrug as a mild opinion.

How scoring and your role summary work

Here is the mechanism, at a level you can follow without reading source code. Each statement contributes positively or negatively to one or more of the nine dimensions. Some statements are written in reverse — agreeing counts against a dimension rather than for it — specifically to catch the tendency to agree with everything without reading closely. For each dimension, your contributing answers are added up and normalized onto a 0–100 scale for readability, after any "not sure" answers for that dimension are excluded rather than counted as a neutral midpoint.

A dimension only appears as a readable score if you answered enough of its statements to make that number meaningful — a minimum coverage threshold has to be met first. If too many dimensions fall short of that threshold because you skipped a lot of statements, you will see a message telling you the profile does not have enough answers yet, rather than a confident-looking number built on almost nothing.

Your dominant, submissive and flexibility scores are then read together to generate one overall summary, not averaged into a single number. The possible families are dominant-leaning, submissive-leaning, switch-flexible or balanced, and mixed or context-dependent — none presented as a permanent identity. A mixed or context-dependent result is a complete, valid outcome, not a failure; that label is reserved for the separate case of too few answers to form a profile at all. Whichever family you land in, the results guide walks through what that specific pattern tends to mean and what to read next.

One pattern is worth explaining on its own, because people expect the opposite: a switch-flexible result never requires your dominant and submissive scores to be numerically equal. Flexibility can be high even when one role scores noticeably higher than the other overall — it is entirely possible to lean dominant most of the time and still register meaningful flexibility, because that flexibility might only show up with one particular partner, or only under certain circumstances. Switching does not have to mean switching constantly, or switching evenly, to be genuinely switching. Am I a Switch? and Dom vs Sub vs Switch both go further into what that actually looks like in practice than a scoring explanation reasonably can.

The research behind this tool, and what it doesn't mean

This assessment was not written in a vacuum. Its terminology, its multidimensional structure and its caution about calling any of this validated all draw on published research into kink, BDSM and consent, summarized in more depth on BDSM psychology. A 2024 paper in the Journal of Sex Research describes and validates the Kink Orientation Scale, a research instrument built to measure kink desire, practice and identity across several dimensions — its existence is part of why a multidimensional approach felt like a reasonable starting point, in general. To be precise about what that paper does and does not lend this tool: this assessment is not the Kink Orientation Scale, does not reuse its items, and does not claim any of its validation. It informed a design philosophy, nothing more.

Formal clinical practice guidelines for working with kink-involved clients, published in 2023 after a multi-year expert review, shaped this tool's refusal to suggest any answer pattern is healthier or more advanced than another. A 2018 clinical review found that people practising BDSM differ minimally from the general population on standard measures of psychopathology, and a larger 2020 review of sixty studies reached a similar conclusion while flagging how narrow and self-selected its own evidence base was — both are why this page treats an interest in any of these nine dimensions as ordinary variation, not something requiring justification. Separate consent research treats mutual, informed, freely given agreement, not any particular role or activity, as what distinguishes BDSM from abuse in the literature. How consent works in BDSM covers that in practice; it is also why this page states plainly, further down, that a high communication or consent-related score here cannot prove you or anyone else actually behaves safely.

Limitations you should know

None of these are hidden in fine print. They are the honest reasons to treat your result as a starting point rather than a conclusion.

• Self-report bias. Your answers reflect how you see yourself in the moment, not how you would necessarily behave under pressure, in real negotiation, or when things go differently than planned.
• Mood and context effects. The same person can answer differently on a stressed weekday than during a calm weekend, and this tool only ever sees one sitting.
• Terminology varies. Words like "structure" or "leading" do not mean identically the same thing to every reader. Your sense of a term and someone else's may not line up.
• Indian cultural context is underrepresented. Almost none of the research this tool draws on was conducted in India, so its framing carries a Western research bias this page cannot correct on its own.
• Fantasy and real life can differ. Finding something appealing to imagine is not the same as wanting it, or being ready for it, outside your own head. See Fantasy vs Real Life for that distinction in full.
• People change. A result today describes today. Interests, comfort levels and preferences shift with experience, relationships and time, and retaking the assessment later can produce a genuinely different, equally valid result.
• The item bank itself is unvalidated. These statements have not been through the psychometric testing — factor analysis, test-retest reliability, external validation — that a genuine research instrument requires.
• Score thresholds are interpretive, not scientific. The cutoffs that decide "leaning" from "mixed" are editorial judgement calls made for a self-reflection tool, not statistically derived boundaries.
• A high communication or consent-related score proves nothing about actual behaviour. Stating that you value clear conversation is not the same as having practised it, and no self-report tool can verify either one.

Privacy and versioning

This page reflects Assessment Version 1.0.0, last reviewed 1 September 2026. Any future change to the item bank, scoring weights, result thresholds, dimension definitions or the copy that describes them will carry a version increment, so a version number you see here always corresponds to the exact test behaviour it describes.