Turning Inquiry-Based Learning into a Portfolio for Japan's Comprehensive Selection: What Gets Evaluated and How to Plan Backward from Grade 10

This is Hamamoto from TIMEWELL.

Roughly one in five university entrants in Japan now enters through comprehensive selection, the admissions route known as sogo-gata senbatsu. According to the latest implementation data published by the Ministry of Education, Culture, Sports, Science and Technology (MEXT) on November 26, 2025, 126,766 students entered university through comprehensive selection, 19.5% of all entrants, and the figure rises to 22.8% at private universities¹. MEXT's admissions guidelines also require that the activity report, a document every comprehensive selection must use in its evaluation, describe research undertaken in the Period for Inquiry-Based Cross-Disciplinary Study, the inquiry class that every Japanese high school student takes². In other words, a system is already running in which ordinary classroom inquiry becomes the central admissions document. In this article, I use only primary government sources to lay out how far comprehensive selection has spread, what separates inquiry that gets evaluated from inquiry that does not, what belongs in a portfolio, and how to plan backward from the first year of high school.

Our core business at TIMEWELL is hands-on AI support for companies, but we also run programs for high school students. In that work I keep hearing the same problem in two different vocabularies. Inquiry coordinators say the results their students worked hard for never connect to admissions. Career guidance teachers say the material for comprehensive selection is still missing right before applications open. The cause is usually the same: nothing was recorded along the way. Three key points first.

• Comprehensive selection and school-recommendation selection together account for 53.6% of entrants. More than half of students now settle their university place through routes decided before general exams, and inquiry records sit at the center of that evaluation.
• What separates outcomes is not polish but three things: the originality of the question, the record of the process, and the depth of verification. Interviews became mandatory from the 2027-entry cycle, and inquiry whose process cannot be narrated collapses there, however handsome the documents.
• Starting a portfolio in the final year is too late. Applications open on or after September 1 by rule, so a logging habit from the first year of high school is what makes the difference.

Comprehensive selection is no longer a special route: the numbers

First, the scale, stated precisely. MEXT's summary of the 2025 admissions cycle shows 126,766 students, 19.5% of all university entrants, entering through comprehensive selection¹. By founder type, the shares are 7.7% at national universities, 6.0% at public universities, and 22.8% at private universities. Adding school-recommendation selection at 34.1% brings the total to 53.6%, meaning more than half of entrants secured their places through selections decided before the general exam season¹. On the supply side, 728 universities, 93.6% of the total, run comprehensive selection¹. Nearly every university in the country has a door that inquiry work can open.

One caveat in the statistics deserves respect. The selection categories were reorganized from the 2025 guidelines onward, so the figures cannot be compared directly with earlier years¹. That is why I have deliberately avoided writing "up X points from last year." Even so, one in five overall, and approaching one in four at private universities, is not a level anyone can call marginal.

Comprehensive selection descends from what used to be called AO admissions. MEXT renamed and rebuilt the category as part of its reform connecting high school and university education, aiming for more multifaceted, holistic evaluation³. The current guidelines define it as "an admissions method that, on the premise of evaluating by viewpoints and methods different from general selection, combines detailed document screening with careful, time-intensive interviews and the like to holistically evaluate and judge an applicant's abilities and aptitude, motivation toward learning, sense of purpose, and so on"². You still hear the old caricatures, a talent-show entrance or a back door around academics. The design philosophy is the opposite. Read the documents closely, talk with the person at length, and look at their motivation and purpose. It is the labor-intensive way to admit students.

For inquiry, the detail that matters most is that the word itself is built into the evaluation axis of Japanese admissions. The guidelines define the second of the "three elements of academic ability" as "the capacity for thinking, judgment, and expression needed to apply knowledge and skills, discover problems on one's own, pursue inquiry toward their solution, and express the results"². Inquiry is not an accessory on the edge of the entrance exam system. It sits inside the definition of what is being measured.

There was also a significant recent change. The 2027-entry guidelines, issued by notice on May 27, 2026, state that comprehensive selection and school-recommendation selection must now include evaluation by interview⁴. Categories that already existed and find introduction difficult have a transition period until the 2029-entry cycle, but the direction is fixed⁴. The background is telling: in the 2026-entry cycle, some universities weighted subject-based tests held before February 1 so heavily that selection effectively turned on academic test scores, against the very purpose of these routes, and the national admissions council issued a formal request to all university presidents to honor the guidelines⁴. The state's position is unambiguous. Comprehensive selection is to remain an admissions route that reads documents and talks to people. Seen from the applicant's side, the change rewards those who have something substantive to say.

What separates inquiry that gets evaluated from inquiry that does not

So what kind of inquiry is strong in comprehensive selection? I see three dividing lines: the originality of the question, the record of the process, and the depth of verification.

The first is originality, or more precisely, connection to the self. The national Course of Study writes into the goals of the inquiry period that students should "find questions out of the connection between real society, real life, and the self, and set their own problems"⁵. The system itself demands a link between the question and the questioner. A project titled "I researched the SDGs" is weak not because the theme is too big, but because no one can see why you are the person asking it. A question that grew out of crop damage on a grandfather's farm and led to agricultural technology holds up under interview probing in a way that a dutifully chosen textbook topic never will. Yoichi Ito, dean of the Faculty of Entrepreneurship at Musashino University, told an audience of high school career guidance teachers that "entrepreneurship education comes down to the question, 'What do you want to do?' Career education is the same"⁶. Strip away the paperwork and that is also the question comprehensive selection is asking.

The second is the record of the process. Recall the phrase "detailed document screening" in the definition². Reviewers are trying to read the path of thinking behind the finished artifact. Yet most students arrive at application season holding nothing but the final presentation deck. What was the first version of the question, where did it stall, whose advice changed it and how? That trajectory is precisely the evidence of having "discovered a problem on one's own and pursued inquiry toward its solution," and without records it cannot be told. To a reviewer, a question rewritten three times, with the drafts to prove it, speaks louder than one tidy conclusion.

The third is depth of verification. A survey of thirty classmates ending in "I felt that..." and a cycle of hypothesis, test, refutation, and revised hypothesis will be scored differently even on the same theme. What the three elements demand is the ability to pursue inquiry toward a solution, not the ability to write a book report². A record of a hypothesis that failed is not evidence of failure. It is evidence that verification actually happened. If anything, an inquiry in which no hypothesis ever failed invites the suspicion that nothing was tested.

Here are the three dividing lines in comparative form.

Dividing line	Inquiry that struggles	Inquiry that gets evaluated
Question	Takes a generic theme as given	Connects the question to the student's own experience and life
Process	Only the final slides survive	Dated records of how the question changed and where it stalled
Verification	Ends with summaries and impressions	Tests hypotheses and folds failed results into the next move

To be clear, a student whose inquiry so far matches the left column does not need to give up on comprehensive selection. Because the dividing lines sit in process and verification, restarting the record and deepening one round of testing can recover a great deal even midway. For students rebuilding their theme, the question-design examples in our 50 AI themes for inquiry-based learning are a useful starting point.

What belongs in the portfolio: working backward from the activity report

Portfolio is a vague word, but for comprehensive selection the required contents can be reverse-engineered from the system with surprising precision. The guidelines state that, given its open-application character, comprehensive selection "must use documents written by the applicant" in evaluation, and they name the activity report, the statement of reasons for seeking admission, and the study plan as those documents². The activity report is expected to describe "research undertaken in the Period for Inquiry-Based Cross-Disciplinary Study, in Rika Tankyu (science inquiry), and the like," and the reference format provides fields for why the research theme was chosen and for its outline and results². The guidelines further encourage universities to make active use of these documents "through presentations and the like"². So what the application season will demand is exactly three things: the reason for the theme, the substance and results of the work, and the material to speak about both aloud.

Working backward from that, the day-to-day portfolio reduces to five kinds of material.

Material	What it proves	How to keep it
Notes on how the question changed	The reason for the theme, deepening of thought	Never delete old versions; date each rewrite of the question
Records of conversations and feedback	Collaborating with diverse people, digesting advice	Three lines: who said what, and what you changed
Prototypes and failed versions	Verification put into practice	Save the versions that did not work, in photos or screenshots
Data and the analysis process	Organizing and analyzing information	Keep raw data and summaries separate; note why you judged as you did
Presentation materials and Q&A records	Summarizing and expressing	Record the questions you received, above all the ones you could not answer

None of this can be reconstructed retroactively in the final year, which is exactly why the routine has to be light. My recommendation is to fix a ten-minute weekly slot for writing down "what changed in my inquiry this week." A paper file works, but a cloud document or slide deck is the practical choice because dates are stamped automatically and everything is searchable. Three lines is enough. Continuity beats everything else.

For students using generative AI in their inquiry, the AI conversation logs are themselves legitimate material. The one discipline that matters is being able to explain, alongside the log, where the AI's work ended and your own thinking began. When an interviewer asks whether the AI did this analysis, the student who can answer "I delegated up to this point, and this judgment was mine" tends to be read as exactly the kind of AI-literate applicant universities want. How to build generative AI into inquiry classes, step by step, is covered in our guide to inquiry-based learning with generative AI.

A working product is the strongest evidence there is

Let me push the materials question one step further. The single most persuasive item a portfolio can contain is a product that actually works. An app, a web service, a device, a product you tried to sell. The reason is simple: in a mountain of documents, a working artifact is a verifiable fact. Claims written in a report have to be taken on trust. An app behind a URL can be tested by the reviewer's own hands.

Mandatory interviews raise this value further. The interview formats named in the guidelines include presentations and oral examination². An exchange about abstract motivations flattens quickly, but with a real artifact on the table, the conversation turns concrete at once. Why this feature, what did users say, where did you get stuck and what did you cut? For a student holding both the process record and the artifact, the interview stops being an interrogation and becomes home ground. In the lecture cited earlier, Yoichi Ito pointed to commercial and agricultural high schools and observed that "when students actually try it, selling their products or crops, their persuasiveness becomes completely different"⁶. Put the prototype into the world, even a rough one. That single step, he argues, changes the weight of everything the student says afterward. Working alongside high school students, I have seen exactly the same thing.

And the threshold for that step has collapsed. With generative AI, a high school student with thin programming experience can reach a working app prototype in a few weeks. The tooling and the development path are laid out in our practical guide to AI development for high school students, but the short version is that buildability is no longer the main barrier. The barrier has moved to whether the thing you build can be derived from your own question. A question raised in inquiry, a record of verification, and a working product at the end of the chain. A portfolio with that spine is formidably strong in comprehensive selection. That is a conviction from practice, not theory.

At TIMEWELL, through WARP for Schools, our program for schools and educational institutions, we walk students from their inquiry question all the way to building a product with their own hands. We have been involved in developing more than 500 people and have worked on a partnership project with the Tokyo Metropolitan Government (WARP ENTRE). For teachers who want to build a school-level design that connects inquiry outcomes to admissions, we are glad to start with a simple exchange of notes.

The backward schedule from grade 10: the deadline is set by the system

Finally, time, which is where collaboration between career guidance and inquiry coordinators pays off most. The guidelines set application intake for comprehensive selection on or after September 1 and results on or after November 1². For the 2027-entry cycle that means September 1, 2026, only about six weeks after this article's publication for current final-year students. The framework has been set the same way in recent years' guidelines², so it can be treated as fixed. Read in reverse: if the core material does not exist by the summer of the final year, the documents can still be written, but there will be nothing inside them.

Working backward from grade 10 (the first year of Japanese high school), a realistic allocation looks like this.

Period	What to do	Portfolio milestone
Year 1	Hunt for the seed of a question; start a ten-minute weekly log	Several dated candidate questions on file
Year 2, first half	Commit to one question; gather information and set hypotheses	Question-change notes and conversation records in motion
Year 2, second half	Verify and prototype; seek feedback from adults outside school	Verification records including failures, prototypes, mid-term presentation
Year 3, spring	Consolidate results; match them against each university's admissions policy	Activity report essentially draftable
Year 3, summer	Finish documents; practice interviews; apply from September 1	The process can be narrated in the student's own words

The most important feature of this table is that the first-year row is the lightest. Year one requires no results, only a logging habit and the seed of a question. Skip the verification and prototyping of late year two, however, and no amount of prose polishing in year three will supply the missing substance. The inquiry period is a compulsory subject allocated 3 to 6 standard credits to every student⁵. The system is designed so that simply accumulating that class time with records attached produces most of the material comprehensive selection needs; it evaporates only because the small habit of recording is missing. Setting a portfolio template at the school level, rather than relying on each student's diligence, is the reliable fix, and that is my honest recommendation to school leaders.

As an aside, this structure aligns with the direction of Japan's national high school reform, which is moving inquiry and cross-disciplinary learning to the center of upper secondary education. I mapped that larger picture in our explainer on the N-E.X.T. High School Initiative. Enriching inquiry and evaluating it in admissions are not separate policies but one continuous current.

Summary

• Comprehensive selection accounts for 19.5% of university entrants overall and 22.8% at private universities. Together with school-recommendation selection, 53.6% of entrants use these earlier-deciding routes, and 728 universities (93.6%) run comprehensive selection.
• The guidelines require activity reports documenting inquiry-period research, and interviews became mandatory from the 2027-entry cycle. Inquiry records are central admissions material.
• The dividing lines are the originality of the question, the record of the process, and the depth of verification. Dated drafts and failed hypotheses count for more than one polished conclusion.
• Keep five materials: question-change notes, conversation logs, prototypes including failures, data with analysis, and presentation materials with Q&A. Ten minutes a week sustains all of it.
• Applications open on or after September 1 each year. Build the logging habit in year one, verify and prototype in year two, and turn to documents and interviews in year three.

Inquiry, in the end, is the practice of growing a question while keeping the record of its growth. That is also precisely the capacity comprehensive selection is trying to see. There is no need to bend inquiry to serve admissions. Do the inquiry seriously and keep the trail, and the range of paths open to a student widens quietly on its own. The three-line log can start in this week's class.

---

References

Related articles

• A Practical Guide to AI Development for High School Students
• 50 AI Themes for Inquiry-Based Learning That Teachers Can Use As-Is
• High School Inquiry-Based Learning with Generative AI: Practical Steps
• What Is the N-E.X.T. High School Initiative? Japan's High School Reform Explained
• What Is AI for Science? How AI Is Changing Scientific Research

Footnotes

1. MEXT, "Summary of the FY2025 Implementation Status of University and Junior College Admissions" (published November 26, 2025) ↩ ↩² ↩³ ↩⁴ ↩⁵

2. MEXT, "FY2027 University Admissions Implementation Guidelines" (Notice No. 318 of the Higher Education Bureau, May 27, 2026, annex) ↩ ↩² ↩³ ↩⁴ ↩⁵ ↩⁶ ↩⁷ ↩⁸ ↩⁹ ↩¹⁰ ↩¹¹

3. MEXT, High School-University Articulation Reform FAQ 3-1-2, "How will the names of admissions categories change?" ↩

4. MEXT, "Notice on the FY2027 University Admissions Implementation Guidelines and Graduate Admissions Implementation Guidelines" (May 27, 2026) ↩ ↩² ↩³

5. MEXT, Comparative Table of the Course of Study for Upper Secondary Schools: Period for Inquiry-Based Cross-Disciplinary Study ↩ ↩²

6. Lecture transcript, "Entrepreneurship Education and Career Education" by Yoichi Ito, 48th Aomori Prefectural Upper Secondary Education Research Conference, Career Guidance Division (FY2024 research bulletin) ↩ ↩²