Guest contribution by Karlis Vilmanis, Founder of Ikigain, an independent project studying structured reflection on purpose and career direction. This piece is editorial analysis, not product promotion; the methods, data and aggregate tables it draws on are public.
A purpose test can be useful when it turns a vague question into something a person can examine. It becomes misleading when a score is treated as a verdict about identity, talent or the future. A new analysis of 4,105 anonymous adults who completed Ikigain’s free 18-item assessment offers a useful distinction: large datasets can describe how a particular group responds to a particular instrument, but they cannot tell us what any one person should do next.
The dataset covers assessments completed from April 1 to July 14, 2026. Each respondent allocated a fixed pool of 18 attention units across four familiar pillars: Passion, Mission, Vocation and Profession. The results are therefore not a measure of how much purpose someone has. They show how attention was divided inside one forced-choice exercise.
The strongest finding is also the most modest. People were rarely balanced across the four pillars. The typical gap between a respondent’s best- and worst-funded pillar was 22.7 percentage points. In other words, most people placed noticeably more weight on some parts of the question than on others. That’s a description of emphasis, not a diagnosis of a weakness.
A correction that changes the story
The analysis also illustrates why transparency matters more than a dramatic headline. An earlier release, based on an older answer-scoring map, reported an 8.1-point gap between Mission and Profession. An internal audit found that the size of that gap depended on the retired scoring map rather than on a stable fact about the respondents. Recomputing the comparison under the current method reduced the gap to 3.1 points. Profession was no longer the clear lowest pillar.
That change doesn’t make the dataset useless. It makes the claim narrower and more credible. The revised analysis is limited to responses collected under the current item set, states what changed, and keeps the more durable result in view: respondents were lopsided in their allocation even when the ranking of the average pillars moved.
The fixed budget changes how correlations should be read
There’s a technical trap in any exercise where four numbers must add up to the same total. If a person gives more units to one pillar, fewer units remain for the other three. Raw correlations will therefore look negative even when there is no meaningful psychological trade-off. A simple chart can mistake arithmetic for a story about people.
To reduce that problem, the analysis treats the four scores as compositional data and uses a centered log-ratio transformation. It then adds a second check: a permutation null test. In 2,000 permutations, the columns are shuffled across respondents so the fixed budget remains but within-person relationships are destroyed. An observed relationship is treated as substantive only when it falls outside the range produced by that mechanical constraint.
That procedure changes the interpretation. Passion and Mission show the strongest genuine trade-off, with a corrected correlation of minus 0.50. Vocation and Profession show a second meaningful tension at minus 0.42. But Mission and Profession move together more than the fixed budget alone would predict. Contribution and livelihood aren’t always opponents; for some people, what the world needs is closely tied to what they can be paid to do.
A label is not a discovered type
The assessment also assigns people to named archetypes. Those labels can help readers talk about a result, but the analysis does not support treating them as natural psychological species. K-means clustering across two to ten groups produced no strong separation; the best silhouette score was 0.28. The response space is better described as a continuum, with some regions more populated than others, than as ten sharply separated types.
This distinction matters for anyone who has ever received a personality result. A label can be a useful prompt: it may help someone notice a preference, question an assumption or choose a small experiment. It should not become a new job title, a reason to reject an opportunity or evidence that a person is incapable of change.
What the dataset can support
The analysis can support careful, practical questions. If Mission is strongly funded and Profession is lighter, a reader might explore which forms of contribution could also be sustainable. If Passion and Mission compete for attention, the reader might test whether a valued activity can serve others in a small, bounded way rather than demanding an immediate career change. If one pillar is consistently low, the useful question isn’t “What is wrong with me?” but “What information would help me understand this result?”
A two-week experiment can provide that information. Someone considering a new field might interview a practitioner, shadow a routine, complete one realistic task, or volunteer for a small project. The purpose of the experiment isn’t to prove a destiny. It’s to replace an imagined future with observations about energy, constraints, learning and fit.
What the dataset cannot support
The sample is self-selected. Respondents arrived at a free online purpose test, so the results describe the audience that this kind of tool reaches, not the general population. The study contains no age, gender, country or income variables, and it measures one occasion rather than change over time. It also studies the popular four-pillar synthesis that uses the word ikigai; it is not a direct measurement of the original Japanese psychological construct or of the validated Ikigai-9 scale.
Those limits rule out several tempting conclusions. The results cannot show that one career is objectively right for a person. They cannot establish that a low pillar causes dissatisfaction, that a type predicts job performance, or that the four pillars should be equal. They cannot replace a conversation with a qualified professional when someone is dealing with mental health, financial or employment risk.
The value of a purpose dataset
The most responsible use of a purpose dataset is not to make identity feel certain. It’s to make uncertainty more workable. A transparent method can show where a pattern is strong, where it’s fragile and what remains unknown. A careful reader can then use the result as a starting point for observation rather than as permission to stop thinking.
That’s the standard the revised analysis sets for itself. It reports a correction instead of hiding it, separates arithmetic from association, tests its labels, publishes aggregate tables and states its blind spots. The practical lesson is equally simple: use a result to design a better question, then let a small real experience supply the next piece of evidence.
Methods and data
The aggregate report, methods note, companion CSV tables and replication contact are available at ikigain.org/research/ikigai-statistics-2026. The dataset contains no individual-level responses. Aggregate cross-tabulations and bootstrap output can be requested directly from Ikigain.
Karlis Vilmanis is the founder of Ikigain, an independent project studying structured reflection on purpose and career direction.
