Methods
Survey Design
The Stanford Japan Barometer is fielded with a nationally representative sample of Japanese adults. Sampling and data collection are conducted by a major Japanese survey firm. Each survey includes embedded experiments alongside standard opinion questions and demographic background questions.
Survey Experiments
Each wave embeds original experiments to identify what drives public attitudes — going beyond descriptive polling.
Conjoint Experiments
Respondents see two hypothetical profiles side by side and choose which they prefer. The profiles are randomly generated from a set of attributes, so we can pinpoint exactly which characteristics shape preferences. Below is an example from the immigration conjoint:
| Applicant A | Applicant B | |
|---|---|---|
| Origin country | Vietnam | Germany |
| Occupation | Nurse | IT engineer |
| Education | University | High school |
| Japanese ability | Fluent | Through interpreter |
| Reason for application | Seek better employment | Escape persecution |
| Prior visits to Japan | Never visited | Multiple tourist visits |
Which applicant would you prefer Japan to admit?
By randomly varying the attributes across many such pairs, we can measure how much each characteristic — nationality, occupation, language ability, and so on — affects whether respondents support admission.
Vignette Experiments
Respondents are randomly assigned to read a short passage framing a policy issue from a particular angle, then answer a question about their views. Because the framing is assigned at random, any difference in responses reflects the effect of the framing itself. Below is an example from the immigration framing vignette:
Accepting immigrants can bring economic benefits to Japan. Immigrants fill labor shortages in sectors such as manufacturing, care, and construction, and contribute to tax revenues that support an aging society.
How much do you support expanding immigration to Japan?
☐ Strongly support ☐ Somewhat support ☐ Somewhat oppose ☐ Strongly oppose
Other respondents receive the same question but with a different passage — framing immigration as a cultural challenge, a security concern, or no framing at all. Comparing responses across groups reveals how the framing shifts opinion.