Each person contributes one pairing: their field studied and the first-ranked field among those closely related to current work. Only responses naming specific fields on both axes are plotted. Responses including Other or None are outside the chart.

Annual income is evaluated using reported brackets. JPY 7 million or more covers brackets from JPY 7–7.99 million through JPY 20 million or more. Upper bounds are exclusive. When an income limit is specified, no-answer and prefer-not-to-answer responses are excluded; without a limit they are included.

Master’s degrees include national, local public, private and overseas graduate schools. Undergraduate degrees and doctorates also include Japan and overseas. Original categories can be selected individually. Education follows the survey response categories.

Exclude physicians and dentists removes occupation code 59. Physicians and dentists share one response category and cannot be separated. This does not identify licenses, medical doctorates, or actual clinical or research practice. People selecting other occupations remain; researchers selecting physicians and dentists are excluded. It does not exclude the entire D field group.

Only four R&D occupations includes basic research (including national research institutes), applied research / advanced development, and the corresponding IT / systems categories: codes 1, 2, 14 and 15. The separate university teacher / researcher occupation category is not included. This filter does not restrict employer size or industry.

1,000 or more employees includes only the 1,000–2,999 and 3,000-or-more categories. No-answer and do-not-know responses are excluded. Industry is unrestricted, so organizations other than companies may be included.

Master’s degree × income ≥ JPY 7 million sets a lower income bound of JPY 7 million, no upper bound, and a master’s degree in Japan or overseas. Turning it off clears both income bounds and education, preserving age, location, field studied and other switches. It stays synchronized with the selectors and is on only when all three conditions match.

Topic filters can be combined with income, education, age, location and field studied. Respondents must meet every active condition. The four R&D occupations do not include physicians and dentists, so enabling both filters gives the same count as the four R&D occupations alone.

Age refers to the survey date. A lower bound of 30 and upper bound of 40 selects ages 30–39; an upper bound of 30 alone selects people under 30. Ages 30 or younger (under 31) and 25 or younger (under 26) can also be selected. These compare different people in 2025, not changes in the same people as they age.

The diagonal share is the number with identical fields divided by the plotted count. Exact field matches do not measure how useful learning is or how successful someone is. Coordinate distances do not measure similarity of knowledge. Explore alongside respondent counts and differences in fields studied.

Location can refer to residence, workplace, final educational institution or high school. All locations includes overseas, distance education, other and no answer. For graduate-school completers, the final institution need not be their undergraduate university.

The colors, arrangement of 298 fields, and logarithmic / linear height scales follow the public visualization. The height maximum is fixed at 900 people. No count or composition adjustment or population expansion is applied.

Source: Cabinet Office e-CSTI human resources analysis and open data, p4ds2025/v1.0. Prepared by Rokugeisha. Research visualization created on October 6, 2026.