Tokyo Gakugei University (東京学芸大学) · Graduate School of Education, Educational Support Collaborative Practice Development Major (Education AI Research Program)
Questions and follow-ups for this admissions interview, grounded in your target and submitted material: your experience, reasons for applying, and study or research plans.
留学
神社
How to use this report
This sample is a graduate-school case; undergraduate reports follow the same structure. Read the likely questions, answer frameworks and follow-ups, then practice aloud using your own experience and submitted material. Frameworks provide an order and key points; complete answer scripts still need your input and practice. Any spoken answer examples are labeled separately. Use the questions for the interviewer and quick sheet for a final review.
A
Questions
Likely questions and answer frameworks
PAGE01
These are mainly likely questions and answer frameworks. Expand each point into your own answer using your target and submitted material, and confirm any missing experience or reasons. Spoken answer examples appear only when provided.
Opening and motivation
01
まず、自己紹介をお願いします。
What they're probingCheck whether the applicant can link existing study to the proposed topic in their own words.
Answer framework (build it in this order)
Give name and information-management background.
Connect the learning-support observation and seminar log visualisation to the question.
Close with the proposed small-unit study, clearly future work.
Likely follow-ups (1)
How would you state your research interest in one sentence?
Based onFictional applicant brief: information management, online learning support, log visualisation
Reasons for applying
02
なぜ大学院で学びたいのですか。なぜ日本で学ぶことを考えていますか。
What they're probingSeparate the reason for graduate study from the personal reason for studying in Japan, which the brief does not supply.
Answer framework (build it in this order)
Explain the need to learn validation design and interpretation of educational data.
Supply a genuine reason for Japan tied to a checked course, environment, or experience.
If that reason is not settled, prepare it before the interview instead of offering broad praise.
Likely follow-ups (1)
What differs from continuing study in your home country?
Based onFictional brief gives research intent, but no personal reason for choosing Japan
03
なぜ東京学芸大学の教育AI研究プログラムを志望するのですか。
What they're probingConnect the actual published program to the proposed work.
Answer framework (build it in this order)
Name the Education AI Research Program as the academic setting.
Relate データ・アナリティクス実践論 to methods and 人間情報学論 to teacher interpretation.
Required field research can inform the setting; data permission is a separate step.
Likely follow-ups (1)
How would your study plan change if those courses differ in the application year?
Based onTokyo Gakugei University official program, course and curriculum information
04
情報管理から教育分野へ進む理由を、ご自身の経験から説明してください。
What they're probingGround the move across disciplines in reported experience.
Answer framework (build it in this order)
State the database, statistics and programming foundations from information management.
Describe seeing different sticking points in the same material during learning support.
Explain what log visualisation showed and why leaving a unit does not prove difficulty.
Likely follow-ups (1)
Which experience most changed your initial question?
Based onFictional applicant brief: study, volunteer experience and seminar work
Study and skills
05
これまでの学修で難しかったことと、どのように学び直したかを教えてください。
What they're probingAsk for a concrete learning process; the brief gives no completed story of overcoming a setback.
Answer framework (build it in this order)
Identify the difficulty of interpreting a log chart as evidence of understanding.
If there was a real learning episode, add its task, sources, revision and result.
Otherwise explain the plan to learn validation design without inventing a past success.
Likely follow-ups (1)
What actual revision of your work shows how you learned?
Based onFictional brief: visualisation assignment and skills; no specific setback supplied
06
機械学習の手法は、どこまで自分で扱えますか。
What they're probingThe panel needs an accurate account of current preparation, not future coursework presented as an already-mastered skill.
What has been studied: foundations in databases, statistics, Python, and SQL.
What has been done: organize and visualise activity logs in a seminar and explain the descriptive finding of units with more exits.
What remains: model choice, validation design, and educational interpretation need systematic study.
The public courses on Python model building and HCI evaluation are directions to study and test for fit, not skills already completed.
Likely follow-ups (2)
Why did the seminar result not establish learning difficulty?
Which skill will be strengthened first in a small exercise?
Based onApplicant material and official course descriptions: information-management foundations, データ・アナリティクス実践論, 人間情報学論
Research plan
07
まず、研究計画の概要を三分程度で説明してください。
What they're probingThe panel needs to hear that the question, the evidence being tested, and the boundary of the conclusion fit together—not a memorised project description.
Scope: one course and a small number of units, kept to a researchable size.
First layer: compare candidate signals such as rewatches and retries with an attainment indicator independent of the logs, checking both signal usefulness and false positives.
Second layer: if teacher collaboration and permission are available, show the unit, behavioral evidence, and material location to inspect; ask separately whether the evidence is readable and why the teacher would revise or not revise.
Boundary: neither the screen, a teacher judgment, nor a material change by itself establishes improved learning outcomes.
Likely follow-ups (2)
What records does each layer require?
Which layer must leave the thesis claim if collaborative data are unavailable?
Based onSample applicant materials: small-unit scope, independent attainment indicator, and separate teacher interpretation/revision-judgement layer
08
この研究計画で、現時点で最も弱い点は何ですか。
What they're probingTest whether the applicant can identify the plan's central dependency.
Answer framework (build it in this order)
The key unknown is permission for course data plus an independent attainment indicator.
With public data only, restrict the claim to testing parts of the analysis procedure.
Teacher interpretation requires separate cooperation and must be omitted if unavailable.
Likely follow-ups (1)
How would the thesis claim narrow without an independent indicator?
Based onFictional future research plan and unconfirmed data, permission and cooperation
09
操作ログのデータは、どこから、どのように入手する予定ですか。
What they're probingThis tests whether the applicant distinguishes a wish to engage with an educational setting from permission to use research data.
Answer framework (build it in this order)
State the fact first: there is no already-permitted course log or partner school.
Required field research may help the applicant understand a setting and discuss compliant research conditions, but it is not a guaranteed route to data.
Logs from the undergraduate volunteer school cannot be reused automatically; they can be considered only after independent contact, consent, and permission.
Without course data, use qualifying public data to test the analysis procedure only, and remove the teacher-interpretation and material-judgement layers from the claim.
Likely follow-ups (2)
Who would need to authorise data use?
What cannot be replaced by public data?
Based onSample applicant materials: required field research, no automatic reuse of old logs, and public data only for procedure validation
10
倫理審査はいつ申請し、承認までどのくらいかかると見込んでいますか。
What they're probingThis checks that ethics, information, consent, anonymisation, and data use come before analysis, without guessing an approval window.
Answer framework (build it in this order)
Order the work: narrow the unit and data, confirm the current review and data-use procedure, then analyze.
Name learner information, consent, anonymisation, and secondary-use boundaries as items to settle through supervision and the applicable procedure.
The turnaround is unconfirmed, so give no number; adjust the sequence to formal instructions and rules.
Without the relevant permission, do not fill the gap with old logs or classroom records.
Likely follow-ups (2)
What needs to be written before review submission?
How would the scope change if permission conditions changed?
Based onSample applicant materials: ethics, consent, anonymisation, and data-use permission remain to be confirmed
11
候補信号の有効性と誤検知、そして教員による根拠の解釈を、どのように分けて評価しますか。
What they're probingThis asks whether signal validity and a teacher's use of evidence can be connected without being collapsed into one score.
Answer framework (build it in this order)
Signal layer: use an independent attainment indicator to assess whether behavioral combinations identify units that need checking, and make false positives visible.
Teacher layer: where collaboration is available, ask teachers to read the unit, evidence, and material location; separately record whether they understood it, proposed a revision, or chose not to revise and why.
Accuracy-type measures serve the signal layer only. A revision is one recorded judgment, not a substitute for the other layer.
Neither layer establishes improved learning outcomes; the claim stops at explainable evidence and judgment.
Likely follow-ups (2)
How would a false positive be shown to a teacher?
Which reasons for not revising need to be recorded?
Based onSample applicant materials: independent indicator, false positives, and separate teacher interpretation/revision reasons
12
教育AI研究プログラムと加藤直樹教授の公開情報を、この研究計画にどのようにつなげますか。
What they're probingThis tests whether public program and professor information can be connected to the design specifically and modestly, rather than turned into general praise or an acceptance assumption.
Answer framework (build it in this order)
Start with program fit: the Education AI Research Program is where the applicant wants to learn data methods alongside questions of educational support.
データ・アナリティクス実践論 maps to preparation in Python model building and validation; 人間情報学論 maps to the HCI and evaluation question of how a teacher reads evidence.
Professor Naoki Kato's public HCI/ICT and learning-support-systems profile, plus the IML public summary, make it necessary to distinguish a pace difference from a difficulty signal.
That is a research-match reason only. Contact, a meeting, current eligibility, and acceptance depend on current official requirements and two-sided confirmation.
Likely follow-ups (2)
How does the IML public project differ from your signal definition?
Why does required field research not equal permission to use data?
Based onOfficial sources: Education AI Research Program, two courses, required field research, Professor Naoki Kato, and the IML public project summary
After graduation and life
13
修了後は、研究で学んだことをどのように生かしたいですか。
What they're probingExplore the next use of graduate learning; the brief does not state a career choice.
Answer framework (build it in this order)
The applicant should identify real options such as further research or educational practice.
Connect the methods learned from signal validation and teacher interpretation to those options.
If undecided, explain the options and decision criteria; do not invent an employer or doctoral admission.
Likely follow-ups (1)
What kind of result would change your next-step choice?
Based onFictional brief confirms research intent only; post-degree destination is absent
Study and adjustment
14
日本語で専門的な議論や授業に参加するために、どんな準備をしていますか。
What they're probingDistinguish documented Japanese preparation from readiness for specialized discussion.
Answer framework (build it in this order)
State the confirmed JLPT N2 and continuing reading of specialist books.
Add actual books, oral practice and a way to clarify a question when needed.
Check current course language and available support in the year's guidance.
Likely follow-ups (1)
How would you clarify a technical question you did not hear clearly?
Based onFictional brief: JLPT N2 and specialist reading; oral performance not supplied
B
Follow-ups
Deep-dive question chains
PAGE02
An interviewer may build on your first answer. Practice separating confirmed facts, your own judgment and your next steps.
01
計画書にある「学習躓きの候補信号」とは、操作ログのどの動きを指しますか。
Why this is askedA rewatch or retry can signal difficulty or careful revision. The panel needs to know these are candidate signals, not direct proof of understanding.
Likely follow-ups (3)
視聴の巻き戻しは、躓きの候補信号ですか、それとも熱心さの表れですか。
その区別を、ログとは独立した到達度指標とどう照合しますか。
誤検知だった場合、どのように扱いますか。
Answer points
Say candidate first: one log action is not a definition of difficulty.
Compare combinations of actions with an attainment indicator independent of the logs, and report false positives as well as supporting cases.
Treat the result as a cue to inspect a unit further, not as a replacement for a judgment about a learner's state.
If an independent indicator cannot be obtained, reduce the claim to testing the analysis procedure.
02
提示を読んだ教員が教材を直さないと判断した場合、その理由をどのように評価しますか。
Why this is askedReading evidence and deciding whether to revise a material are different actions; a decision not to revise can also have a reasoned basis.
Likely follow-ups (3)
根拠を読めたことと、改訂する判断はどう分けますか。
単元、行動の根拠、教材の見直し候補をどのように示しますか。
この評価を学習成果の改善と呼ばない理由は何ですか。
Answer points
Record separately whether the teacher can read the behavioral evidence, locate the material to inspect, and explain a decision to revise or not revise.
A decision not to revise is still a judgment; do not reduce the result to counts of edits.
The screen presents unit, evidence, and a candidate location so interpretability can be examined; it does not promise a learning outcome.
With a small collaboration, the result is design learning, not an effect size or a general material-effect estimate.
03
必修のフィールド研究を含め、二年間で研究をどのように進めますか。
Why this is askedThe plan must explain the dependencies among required field research, permission, and the two evaluations without pretending the timetable has already been coordinated.
Likely follow-ups (2)
協力先や許可が得られない場合、何を検証範囲から外しますか。
公開データで確認できることと、確認できないことは何ですか。
Answer points
In year one, compare literature, define terms and indicators, and include the required field research in the study plan.
After narrowing the object, confirm ethics, consent, anonymisation, and data use through the actual procedure; do not forecast approval or collaboration dates.
In year two, analyze and prototype if permission is in place, then evaluate teacher reading and reasons, with timing adjusted to real conditions.
Without course data, public data test the analysis procedure only; they replace neither required field research nor the teacher-judgement claim.
C
Faculty
Questions from the professor's perspective
PAGE03
These practice questions draw on checked university and laboratory sources listed under Evidence and checks. They are not the professor’s own words or predictions of actual interview questions.
IntentThe IML public project discusses a mismatch with lesson pace. This checks that pace mismatch is not substituted for learning difficulty and that unread details are not invented.
The shared point is seeking clues in material-related activity records that merit further inspection.
This study compares candidate difficulty signals with an independent attainment indicator and makes false positives explicit; a lesson-pace mismatch is not itself difficulty.
Rely only on the public summary, without claiming the IML method, results, a principal investigator's view, or originality for this study.
Follow-up drill:Which object, indicator, and evaluation details need checking in the full text before making the comparison?
SourceOfficial IML public project summary and Tokyo Gakugei University's official profile for Professor Naoki Kato; no contact, acceptance, or individual research detail is claimed.
IntentUsing the publicly stated HCI/ICT research match, this asks whether teacher comprehension and a revision decision can be observed separately.
First assess whether teachers can read the unit, behavioral evidence, and material location; do not fold readability into revision behavior.
Then retain reasons for a proposed revision, a decision not to revise, or an inability to decide.
Do not call either a teacher response or the existence of the screen an improvement in learning outcomes.
Follow-up drill:If a teacher reads the evidence but decides not to revise, how will that judgment be recorded?
SourceTokyo Gakugei University's official profile for Professor Naoki Kato: HCI/ICT and learning support systems. Research match does not mean supervision, collaboration, or acceptance.
03
加藤教授の研究室を志望する前に、公開情報から分かることと、面談で確認したいことをどう分けますか。
IntentSeparate a research match from the supervision and application conditions that need discussion.
The university profile supports HCI/ICT and learning support systems as research areas; connect these carefully to the proposed teacher-readable evidence.
The lab page has carried a preliminary-meeting notice for prospective graduate students; recheck its current scope and booking process.
At a meeting, ask about the plan's fit with supervision, data conditions and preparation before enrollment, without assuming acceptance or collaboration.
Follow-up drill:Which two materials can you explain at a meeting, and which unconfirmed condition would you ask about?
SourceTokyo Gakugei University profile for Professor Kato and official lab page; recheck the current status of the 2026-05-17 preliminary-meeting notice.
D
Region
Region & daily-life questions
PAGE04
Explain which living, funding and arrival arrangements are settled and which still need checking.
01
二年間の学費と生活費について、奨学金の申請予定、家族からの支援、貯蓄をどう組み合わせますか。
Why this is askedFunding is an applicant fact that remains unconfirmed. The exercise is to separate confirmed and unconfirmed items, not to invent figures or a school's support rules.
Answer each source with three parts: source of funds, personally confirmed evidence or status, and the remaining unconfirmed gap.
List scholarships, family support, and savings only where the applicant can substantiate the plan or fact; otherwise say it remains unconfirmed.
Explain that the plan will be updated after costs and time constraints are confirmed, rather than filling the answer with a guessed total.
Follow-up drill:If one source is still unconfirmed, how will you explain that without inventing a total?
02
住まい、来日・転居の時期、通学方法について、確認済みのこととこれから決めることをどう説明しますか。
Why this is askedCurrent residence, arrival or move, and housing are unconfirmed. This is not a school commute rule; it practices separating facts, next checks, and the notice conditions.
Separate what is already known about location or timing from housing, move date, and commute method that still need a decision.
Choose after checking the examinee notice, enrollment date, personal budget, and actual route—not against an invented fixed commute threshold.
If no plan can yet be stated, say when it will be confirmed with the applicant, family, or relevant office instead of inventing a date.
Follow-up drill:If the move date is unconfirmed, who will you ask next and what will you confirm?
03
入学後の授業やフィールド研究に参加するため、生活と学修の予定をどう整えますか。
Why this is askedCourses, field research and discussion in Japanese need a practical schedule. Residence and timing remain unconfirmed.
Start with the confirmed N2 and specialist reading, then add real oral and seminar-discussion practice.
Check the year's course and required field-research schedule before fitting travel, part-time work or other personal commitments around it.
If residence or funding remains unsettled, state the known arrangement and the next check rather than a guessed plan.
Follow-up drill:If field research conflicts with another commitment, which requirement would you check and how would you adjust?
Why ask thisIt connects required field research with permission conditions and shows that field activity does not automatically grant data access.
03
入学までの期間に統計や機械学習を学び直すとしたら、どの領域を優先しておくべきでしょうか。
Why ask thisThe course names are already checked against official information; this asks for a judgment about what to strengthen before enrollment.
F
Quick Sheet
Quick prep sheet
PAGE06
Strategy
Connect motivation to real learning-support and data work; test candidate signals in a few units, then assess how teachers read evidence if cooperation is available.
QUICK PREP SHEET
Quick prep sheet
Run through this on the way — flow, self-intro, strengths to land, questions to ask back, and the day-of checklist.
StrategyConnect motivation to real learning-support and data work; test candidate signals in a few units, then assess how teachers read evidence if cooperation is available.
Interview flow at a glance
Arrival and greetingFollow the noticeUse the location, entry method, and instructions on the examinee notice or contact.
Self-introductionPractice for about 60 secondsConnect name, preparation, the field observation, and the research boundary in one answer.
Research-plan explanationFollow the noticeLimit the scope first, then explain signal testing and teacher judgment as separate layers.
Follow-ups on the planFollow the noticeName unconfirmed permission, collaboration, and procedures before describing the conditional plan.
Academic and language backgroundFollow the noticeKeep the visualisation experience distinct from methods still to be learned.
Questions for the panelFollow the noticeAsk about current ethics, required field research, and permission processes rather than repeat public information.
CloseFollow the noticeThank the panel and follow the in-person or online closing instructions.
Open with the name and ‘I studied information management as an undergraduate,’ then place the existing preparation.
Use the difference seen in learning support as the motivation, without treating the observation as a causal finding.
Use the log-visualisation seminar project as preparation to organize and explain data, while saying that drop-out alone does not establish difficulty.
Describe the small-unit study, independent attainment comparison, and readable evidence for teachers as future work, not a promise of outcomes.
Strengths you must land
There is hands-on preparation in data organization and visualisation, without calling a drop-out signal difficulty itself.
EvidenceThird-year seminar: visualised learning-platform activity logs, identified units with more exits, and explained the figures; the exercise did not establish why students exited.
When to useUse this when asked what can already be done, to bound preparation against methods still to learn.
The question came from a field observation, while retaining the limit of that observation.
EvidenceSecond-year online learning-support volunteering: students using the same material got stuck in different places, and the teaching side could not easily see the difference.
When to useUse this when asked why educational technology or why records of the learning process matter.
Evaluation separates a signal layer from a teacher-judgement layer; neither accuracy nor revision alone replaces them.
EvidenceConfirmed future intention: compare log signals with an independent attainment indicator for false positives; with teacher collaboration, separately ask whether the evidence is readable and why a material would or would not be revised.
When to useUse this when asked about research value, evaluation, or the role of the proposed screen.
Cautions
When a partner, data permission, or supervisory condition is unconfirmed, state the condition and the scope; do not present it as settled collaboration.
Do not invent an accuracy figure. Start with how the independent indicator, false positives, and teacher judgment will each be evaluated.
Do not translate a screen, a teacher reading it, or a material change into improved learning outcomes.
If the Japanese stalls, pause and restate the point in shorter Japanese. Use another language or accommodation only when the notice permits it.
When mentioning Professor Kato or an IML project, stay within the public summary actually read; do not claim an algorithm, result, or originality from an unread full text.
The interview format, arrival or connection time, documents, and online setup are set by the examinee notice or current instructions.
Day-of checklist
Examinee notice, photo ID, and all materials the notice requests.
The submitted research plan and a one-page seminar-visualisation summary that does not contradict it.
Say the self-introduction, the two evaluation layers, and the reduced scope without course data aloud once each.
Confirm format, start time, location or link, and contact route from the notice.
For an in-person interview, leave sensible travel buffer based on the confirmed route; online, test connection, audio, camera, and backup contact.
Keep only official sources or notes whose source and limits can be explained.
Check that every claim about funding, housing, collaboration, and procedure is one the applicant has personally confirmed.
G
Evidence
Evidence and checks
PAGE07
This is a curated sample. Li Ming’s background and research intentions are fictional inputs; the university, course and faculty information comes from the official sources below. Naoki Kato is the selected research-fit target, not a confirmed contact, supervisor or data provider. These are practice questions, not university exam questions or the professor’s own views.
University and research sources
Confirm the requirements for your actual application year separately.
Identifies the Education AI program and Naoki Kato’s HCI and ICT-supported teaching and learning research. It does not establish willingness to supervise this applicant.
The Python modeling course and the HCI and evaluation course address two different learning needs: analytical methods and evaluation of the teacher-facing interface. Open next-year course listings before applying to check availability.
The two-year plan must include required field research. Testing an analysis workflow on public data neither replaces that requirement nor establishes a partner school for the applicant’s own study.
The lab describes a student project using digital-textbook interaction logs to detect divergence from lesson progress. This supports a comparison with comprehension difficulty, not a claim of applicant authorship or novelty. Check the lab’s preliminary-meeting notice and distinguish its non-degree research-student notice from master’s admissions.
Personalization map (why this content is yours)
Shows the material referenced in each section and why it strengthens the content.
Self-introduction scriptFictional applicant Li Ming; information management, learning support, and log visualisation — It shows preparation while retaining the boundary that a log alone does not determine learning difficulty.
Mock Q&AConfirmed future intention: small units, independent indicator, teacher reading and reasons — It rehearses signal validity, false positives, and teacher judgment as separate layers.
Professor-view follow-upsOfficial public information on Professor Naoki Kato and IML — It makes research match traceable without inventing acceptance, collaboration, or personal views.
Study and living arrangementsN2, specialist reading and required field research; funding, housing and schedule remain unconfirmed — It connects academic preparation to practical arrangements without inventing a budget or living plan.
Additional information needed
Answering these will improve the next run's quality.
Items to verify (fact checks)
⚠The Education AI Research Program, データ・アナリティクス実践論, 人間情報学論, and required field research were checked in official sources. Recheck the application-year offering, enrollment, and eligibility conditions in current guidance.
⚠Professor Kato's official page has a note dated 2026-05-17 saying that prospective graduate students should first have a meeting. Confirm its current status, scope, and booking method. Do not expand the 2025-10-16 notice about non-degree research students into a claim that the whole master's program is closed.
⚠The IML public summary supports only the description of digital textbooks, material-operation logs, and mismatch with lesson pace. Without reading the full work, do not infer method, evaluation results, or an individual's contribution.
⚠Ethics review, consent, anonymisation, data use, and collaboration conditions require confirmation for the actual research object and current procedure. This sample does not estimate approval timing.
