The Ivy Institute | College Admissions Research | November 2026
Executive Summary
This study asks a deceptively simple question: when a student’s college-application activity list grows during advising, what actually changed? Did the student remember an experience that had been missing, begin something new, split one broad commitment into several components, or merely describe the same work more precisely? The answer matters because activity-list advice often treats every later addition as proof that students routinely “forget” important experiences. That conclusion is attractive, but it is not automatically supported by a before-and-after comparison.
The Ivy Institute reviewed authorized internal materials, data, and information for seven students whose earlier and later activity accounts could be paired with sufficient confidence. The records covered work conducted from 2025 through 2026 and included students in the later stages of high school. To protect students, this public report does not reproduce private writing, names, schools, organizations, or distinctive combinations of activities. The unit of analysis was a distinct activity entry in each student’s principal later inventory. Duplicate drafts, honors, and optional alternatives outside that principal inventory were excluded.
Across the seven paired cases, the later inventories contained 84 distinct entries. Fifty-four entries, or 64.3%, had a clear counterpart in the earlier account. They were not additions in the substantive sense; they were the same underlying experiences with altered titles, combined roles, expanded descriptions, or more specific results. Thirty entries, or 35.7%, lacked a clear earlier counterpart. Yet chronology changed the interpretation of those thirty. Thirteen entries, or 15.5% of the full later inventory, were demonstrably new experiences begun after the earlier account. Ten entries, or 11.9%, could not be dated precisely enough to determine whether they predated the earlier account. Only seven entries, or 8.3%, met the study’s strict definition of a previously unlisted experience: the later record established that the experience had already begun before the earlier account, but the earlier account did not contain it.
This means that fewer than one quarter of all apparent additions were verifiably pre-existing omissions: 7 of 30, or 23.3%. The majority of apparent additions were either genuinely new or chronologically uncertain. Put differently, a simple document comparison would have overstated the verified omission count by more than fourfold. The finding does not imply that students rarely overlook experiences. It shows instead that a responsible measurement method must distinguish discovery from development and must preserve uncertainty when dates are incomplete.
The seven verified previously unlisted entries were heterogeneous. They included independent creative work, community engagement, service, work, research, and student representation. The pattern therefore did not point to a single “forgotten activity type.” What connected the entries was their weak fit with the conventional mental model of an extracurricular résumé. Some happened outside a named school club. Some had no obvious award. Some blended personal interest with service or intellectual work. Some were embedded in ordinary routines rather than announced as formal programs. The study cannot establish why any particular entry was absent, because the records rarely documented the student’s reason for omission. “Forgotten” is therefore used in the title as a familiar public phrase, not as a claim about memory.
The matched entries also carried an important lesson. Advising frequently changed representation without changing the underlying experience. A general membership could become a role-based entry; related responsibilities could be merged; an award could be reinterpreted as evidence of sustained activity; and a one-line interest could be considered as a fuller experience. Those transformations may improve completeness and clarity, but this study did not test admission outcomes and cannot show that any wording, addition, or ordering increased a student’s chance of admission.
The practical conclusion is a two-pass inventory process. First, families should reconstruct experience broadly: paid work, recurring family or community responsibilities, independent learning, creative practice, hobbies, informal teaching, research, religious or cultural participation, athletics, school organizations, and summer commitments. Second, they should verify dates and decide what belongs in the limited application space. The goal is not to reach a target number. Common App itself advises applicants to focus on meaningful experiences and permits up to ten activities; it explicitly names family responsibilities, jobs, internships, volunteering, hobbies, and sports as possible content. A comprehensive inventory is a discovery tool, not a mandate to submit everything.
The study is intentionally modest. Seven students are not representative of The Ivy Institute’s full student population or of college applicants nationally. The sample was selected because pairable chronological evidence existed, which may favor more documented and intensively advised cases. The analysis evaluates reporting transitions, not the intrinsic value of activities or the quality of counseling. The Ivy Institute conducted and funded the research and is publishing it on its own website; it has a commercial interest in admissions advising. These limits should travel with every use of the findings.
Research Questions
The study was designed around five questions.
First, how many entries in students’ later activity inventories had a recognizable counterpart in their earlier accounts? This question measures continuity. It captures the large amount of revision work that changes language, emphasis, or organization while preserving the same underlying experience.
Second, among entries without an earlier counterpart, how many were demonstrably new? A student who begins a research project, paid job, volunteer role, team, or summer program after the baseline has not “forgotten” that experience. The activity did not yet exist at the earlier observation. Treating it as an omission would confuse personal development with incomplete reporting.
Third, how many later-only entries clearly began before the earlier account? These are the study’s verified previously unlisted experiences. Even here, “unlisted” does not necessarily mean “forgotten.” The earlier account may have been built for a different audience, limited to formal achievements, prepared before a student understood the breadth of the application question, or intentionally selective.
Fourth, what forms did the verified previously unlisted experiences take? The study sought recurring patterns without exposing individual students. It therefore grouped the evidence conceptually and reports only broad descriptions. With just seven students, the report avoids detailed cross-tabulations that could enable re-identification.
Fifth, what process can families use to build a complete and honest inventory without turning the exercise into résumé inflation? Completeness and selectivity are different stages. A family can first identify all meaningful commitments and then choose a smaller set for a particular application. The method must also guard against exaggeration, double-counting, unsupported metrics, and the temptation to transform ordinary involvement into grandiose claims.
The study did not ask whether an activity was “good enough,” whether one category was preferred by admission officers, or whether advising caused an admission decision. It did not score students, estimate admission probabilities, or compare clients with non-clients. No outcome claim should be inferred from the descriptive findings.
Why Activity Inventories Are Difficult
An activity list looks like a form, but creating it is an act of classification. Adolescents do not experience their lives as ten neatly separated entries. A student may help in a family business, translate for relatives, pursue a technical hobby, teach a younger child, train for a sport, write online, participate in a cultural community, and conduct school-based research without using the word “activity” for all of those commitments. The application asks the student to convert lived time into categories, titles, dates, hours, weeks, organizational labels, and short descriptions.
That translation creates several predictable sources of loss.
The first is the institution bias. Students often remember experiences attached to a recognizable institution: a school club, varsity team, named nonprofit, formal internship, or selective program. They may overlook work done independently or within a household, neighborhood, faith community, online community, or informal peer network. This is not necessarily carelessness. School culture commonly uses “extracurricular” as shorthand for organized clubs and competitions.
The second is prestige filtering. Students may decide that an experience is too ordinary to mention because it produced no title or award. Paid work, recurring caregiving, translation, creative practice, personal projects, routine volunteering, and self-directed learning can disappear under this filter. Yet Common App’s own guidance defines the relevant space broadly. The official first-year guide says applicants may share work, hobbies, clubs, community engagement, responsibilities, and circumstances. Its activities worksheet asks students to consider family responsibilities, jobs, internships, volunteer work, hobbies, and sports.
The third is résumé inheritance. Many students begin with a document created for a summer program, award, employment application, or school requirement. That earlier document may emphasize honors, leadership, and technical skills while excluding personal commitments. When it becomes the starting point for a college application, its original purpose silently shapes what appears to count.
The fourth is role compression. One organization can contain several forms of work: participation, leadership, event planning, instruction, fundraising, research, communications, and service. Students may list the organization but omit the most meaningful function, or they may overcorrect and split a single coherent commitment into multiple entries. Both choices can distort the record.
The fifth is time. High-school lives change rapidly. A list written in spring may be genuinely obsolete by fall. Summer programs begin, jobs change, projects are completed, titles shift, and responsibilities expand. Without dates, a later-only item looks like an omission even when it is simply new.
The sixth is the application’s compression. Common App’s published activity worksheet allows up to ten activities and, in the version described by that resource, provides brief fields for position, organization, and details. Limited space encourages selection and abbreviation. A student can maintain a complete personal inventory while submitting only the experiences most useful for a particular application. The existence of an omitted item in the submitted list would not by itself prove an oversight.
Finally, the category has equity implications. Common App’s 2023 research summary analyzed nearly six million activity descriptions from about 860,000 applicants and documented large differences in the number of activities reported across socioeconomic, first-generation, school, and racial or ethnic groups. The researchers cautioned that financial, cultural, logistical, school, and college-preparation differences shape both participation and reporting. They also urged attention to quality and context rather than raw quantity. That national evidence does not validate The Ivy Institute’s internal counts, but it warns against treating a longer activity list as a neutral measure of merit.
For all of these reasons, the key analytic question is not merely whether a line appeared later. It is whether the underlying experience existed earlier, whether it was already represented in another form, and whether there is enough evidence to say.
Study Design And Definitions
3.1 Study setting and company role
The research was conducted by The Ivy Institute using authorized internal materials, data, and information created or maintained through its advising work. The relevant period was 2025 through 2026. The Ivy Institute is a commercial college-admissions advising organization. It selected the question, conducted the review, funded the work, and is publishing the report. That relationship creates an institutional interest in demonstrating the usefulness of structured advising. The design therefore prioritizes reproducible definitions, conservative classification, explicit uncertainty, and separation of descriptive findings from claims about effectiveness.
3.2 Population and analytic sample
The target population was students for whom an earlier activity account and a later application-oriented activity inventory might both exist. The analytic sample was not every Ivy Institute student. A case was eligible only when the research team could establish that the two accounts belonged to the same student, establish their order with reasonable confidence, identify a principal later inventory, and compare underlying experiences rather than only surface wording.
Seven students met the inclusion standard. The paired material spanned 2025 and 2026. All were in the high-school application-planning period, but the exact grade and interval varied. The sample is best understood as a documented paired case series. It is not a random sample, a census of clients, or a matched comparison group.
Cases were excluded when chronology was unclear, the earlier account was unavailable, the later account consisted only of brainstorming fragments, the student identity could not be paired confidently, or the later material did not contain a sufficiently complete principal inventory. Exclusion protects validity, but it also creates selection bias: students with cleaner documentation were more likely to enter the study.
3.3 Unit of analysis
The primary unit was the distinct activity entry in the student’s principal later inventory. The analysis identified 84 entries across seven students. It did not treat every paragraph, draft, edit, or description as a separate observation. Duplicate versions were collapsed. Honors were excluded unless the later inventory explicitly treated the underlying participation as an activity. Optional alternatives outside the principal inventory were not counted in the main total. This rule prevents a student with many revisions from carrying more weight simply because more text existed.
The student remained the unit for sample size and privacy. The report does not publish student-by-student tables. Item counts describe reporting transitions, not independent people. Because multiple entries come from each student, conventional statistical tests that assume independent observations would be inappropriate. The study reports counts and percentages only.
3.4 Matching rule
A later entry was coded as “represented or reframed” when the earlier account contained the same underlying experience. Exact names were not required. A match could be established through a consistent organization, role, project, topic, or set of responsibilities. Changes in title, order, metrics, description, or scope did not create a new activity if the underlying commitment was continuous.
For example, a baseline entry described as a general robotics role and a later entry described as design leadership on the same team were treated as one underlying experience. A later entry combining two stages of the same leadership progression also remained matched. Conversely, participation in a new summer research program was not matched merely because the earlier account contained another research experience in the same broad field.
3.5 Classification of later-only entries
When no baseline counterpart existed, the item entered one of three chronology categories.
“Previously unlisted” required affirmative evidence that participation began before the earlier account. A grade range, dated start, or unambiguous duration could establish this. The absence of a date was not resolved by intuition.
“Genuinely new” required evidence that the experience began after the earlier account. This category includes projects, programs, jobs, and roles that arose during the normal development of the student’s profile.
“Indeterminate” applied when the records did not establish which account came first relative to the activity’s onset. Indeterminate is not a residual judgment about quality. It is an honest statement that the study cannot tell whether the student omitted the experience earlier or began it later.
This classification deliberately avoids the word “forgotten.” Memory is a psychological state and was rarely documented. An unlisted activity could have been forgotten, intentionally excluded, considered irrelevant, compressed into another line, or omitted because the earlier account served a different purpose.
3.6 Coding and verification
The analysis proceeded in four passes. First, the research team inventoried potential pairs and removed cases without adequate chronological or identity support. Second, it normalized each principal later inventory into distinct underlying experiences. Third, it searched the earlier account for direct or functional counterparts. Fourth, it reviewed onset evidence for every later-only entry.
The public calculations are reproducible from the aggregate appendix. The restricted source ledger retains the exact source mapping and item-level decisions. That ledger is separate from the public package because even de-identified combinations of activities can identify a student. The public report provides the decision rules, totals, denominators, and rounding conventions without exposing private source material.
Percentages use 84 later entries as the denominator unless stated otherwise. The apparent-addition denominator is 30. Percentages are rounded to one decimal place. No missing item was imputed. No entry was assigned to “previously unlisted” unless the chronology standard was satisfied.
3.7 Privacy and suppression
The analysis removed names and does not quote students. It does not publish the exact combination of activities associated with any student. Broad examples in the practical guide are synthetic composites, not disguised case narratives. Small category-by-student cells are suppressed. Public data contain only aggregate counts needed to reproduce the displayed figures.
3.8 Outcome boundary
The study measures changes in representation. It does not measure admission-office reactions, admissions results, student satisfaction, hours of advising, or causal effects. A clearer inventory may be practically useful, but the present design cannot show that it changed an admission decision. No later activity was treated as inherently stronger than an earlier one.
The Paired Sample
The final sample included seven paired students and 84 later entries, an average of 12 entries per student before application-specific selection. The average is descriptive only. Some principal inventories included more than the ten spaces available in the first-year Common App activity section because they were working inventories designed to support later selection. The study did not interpret a longer working list as a recommendation to submit more items.
The records varied in interval. In some cases, the earlier and later accounts were separated by weeks; in others, by a school year or more. Longer intervals naturally create more opportunity for genuinely new activity. This is one reason the main analysis uses chronology categories rather than treating every new-to-list item as an omission.
The seven pairs also varied in the purpose and completeness of the earlier account. Some were relatively detailed and already contained leadership, research, work, service, athletics, and honors. Others were narrower. A paired design can identify reporting change, but it cannot fully standardize the purpose for which each earlier account was created. The report therefore calls the earlier material a baseline account rather than a “complete résumé.”
Coverage was adequate for the central comparison but not for every desirable variable. The study could identify principal later entries and earlier counterparts for all seven students. It could not always verify exact start month, weekly hours, whether an entry was ultimately submitted to a college, or why the student had not listed an experience earlier. Those gaps are reflected in the indeterminate category and in the limits on interpretation.
The sample’s small size permits close review but narrows generalizability. The research should be read as an internally documented case series that tests a measurement method and surfaces plausible patterns. It should not be read as an estimate that 8.3% of all applicants’ activities are forgotten or that a typical family will discover a particular number of experiences.
Main Findings
Figure 1 · 84 later entries
What the later entries represented
| Classification | Count | Share | Meaning |
|---|---|---|---|
| Represented or reframed | 54 | 64.3% | Same underlying experience appeared earlier |
| Genuinely new | 13 | 15.5% | Began after the earlier account |
| Chronology indeterminate | 10 | 11.9% | Onset could not be established |
| Previously unlisted | 7 | 8.3% | Demonstrably began before the earlier account |
5.1 Most later entries were continuations, not discoveries
Of the 84 entries in the principal later inventories, 54 had a clear baseline counterpart. That is 64.3% of the total. The majority of the activity-list work visible in the paired records therefore concerned representation: choosing a role title, consolidating related work, making responsibilities concrete, clarifying progression, or connecting an outcome to the student’s contribution.
This matters because a surface comparison can exaggerate change. If an earlier account says “member” and a later account says “workshop leader and outreach coordinator,” the wording has changed substantially. Yet if both refer to the same continuous experience, the later line is not a newly discovered activity. It is a refined account.
The matched category also shows why document-diff methods are insufficient. Automated text comparison would be tempted to call renamed organizations, combined roles, and dramatically expanded descriptions “new.” The study instead matched underlying commitments. This reduced the false-addition problem and produced a more conservative estimate.
5.2 Thirty entries looked new, but chronology divided them
Thirty of 84 entries, or 35.7%, had no clear baseline counterpart. These were the apparent additions. A naïve before-and-after study could stop here and conclude that advising added thirty activities. The chronological review produced a different result.
Thirteen of the thirty apparent additions were genuinely new. These represented 43.3% of apparent additions and 15.5% of all later entries. They included experiences that began after the baseline, such as new research, summer programs, service projects, internships, or senior-year work. Their appearance is evidence that students continued to grow, not evidence that the baseline failed.
Ten apparent additions remained indeterminate. They represented 33.3% of apparent additions and 11.9% of all later entries. The material established that the entry appeared later but did not establish whether the experience itself predated the baseline. The correct result is uncertainty, not a forced assignment.
Seven apparent additions were verified previously unlisted experiences. They represented 23.3% of apparent additions and 8.3% of all later entries. For each of these seven, the later material established that participation had begun before the earlier account. This is the narrowest defensible measure of experiences that were available to be listed earlier but were not.
5.3 The fourfold-overstatement problem
The difference between thirty apparent additions and seven verified previously unlisted experiences is not a minor technical adjustment. A method that counted every later-only entry as “forgotten” would have reported 30 instead of 7—more than four times the supported number.
That overstatement would combine three different phenomena: development, missing chronology, and prior omission. It could also make advising appear more transformative than the evidence permits. The study’s most consequential finding is therefore methodological: in longitudinal application research, a later appearance is not an onset date.
5.4 Previously unlisted experiences were broad, not uniform
The seven verified entries did not fall into one narrow bucket. Collectively they included independent creative practice, community or nonprofit work, research, environmental or civic engagement, and student representation. Some were formal, some informal, and some self-directed. The common thread was not an absence of effort. It was that the experience could be overlooked when a student’s initial mental model favored named school organizations, awards, or conventional leadership titles.
Because the sample is small, this report does not publish counts for each of those subtypes. A detailed table could reveal a student through a distinctive combination. The evidence supports a broad qualitative conclusion: a reconstruction process should ask about contexts, responsibilities, and repeated actions, not only clubs.
5.5 New experiences were a substantial part of the later list
The thirteen genuinely new entries deserve emphasis. High-school students are not static units observed twice. They pursue new opportunities, finish old projects, accept jobs, enter programs, build teams, and take on senior-year roles. When an advising relationship spans months, some “growth” in the activity list should be expected even if the earlier inventory was perfect.
For evaluation research, this creates a time-varying exposure problem. Advising and student development occur simultaneously. A before-and-after design without an onset check cannot attribute all change to the advising process. The present study does not attempt that attribution.
5.6 Uncertainty was common enough to matter
Ten of 30 apparent additions could not be dated confidently. Excluding uncertainty would make the results look cleaner but less true. The indeterminate share signals a practical recordkeeping lesson: students should maintain start dates, grade levels, and brief notes about major role changes as experiences occur. Contemporary documentation is more reliable than reconstructing a multi-year timeline under deadline pressure.
For researchers, the indeterminate category also prevents hidden imputation. It would have been easy to assume that a longstanding-sounding hobby or polished role must predate the baseline. The study rejected that approach. Plausibility is not proof.
5.7 Working inventories were broader than submission spaces
The principal later inventories averaged twelve entries, while Common App guidance permits up to ten in its first-year activities section. This difference illustrates the purpose of a working inventory. Discovery should be expansive; submission should be selective. A student may need a list longer than the final form to compare significance, duration, relevance, and redundancy.
The study did not verify that all 84 entries were ultimately submitted. It therefore reports them as later inventory entries, not application submissions. This distinction is essential. A brainstorming list, an advising inventory, and a submitted application are different analytic objects.
What The Verified Omissions Had In Common
6.1 Weak institutional packaging
Several verified previously unlisted experiences did not resemble a standard school-club line. An independent creative practice, for example, may have no supervisor, official organization, election, certificate, or competition. Yet it may involve sustained production, feedback, audience, collaboration, and skill. The absence of institutional packaging can make real work less visible to the student who performed it.
The same issue applies to neighborhood and community engagement. A student may regularly help, teach, coordinate, research, or create without joining a named organization. A reconstruction interview that asks only “What clubs are you in?” is structurally unable to recover these experiences.
6.2 Everyday commitments can feel too ordinary
Activities embedded in routine life often lack a dramatic beginning. A student may gradually become the person who explains technology, edits peers’ work, helps a community group, cares for younger relatives, or takes responsibility for a recurring local need. Because the role emerged incrementally, the student may not label it as an achievement.
The public title of this study uses the word “forget,” but ordinariness may be more accurate than memory failure. Students remember the experience; they simply do not initially classify it as application material.
6.3 Independent work crosses category boundaries
Self-directed work can be difficult to place. Is sustained online writing an art activity, a hobby, an intellectual project, or community participation? Is independent research a class extension, career exploration, or science activity? Is environmental data collection service or research? The application may require one category, but the lived experience spans several.
When students cannot find a clean label, they may omit the experience rather than tolerate an imperfect category. A better process names the action first and chooses the form category second.
6.4 Service and representation may be hidden inside another identity
Students sometimes separate “leadership” from “service” too sharply. They may list an elected position but omit the ongoing peer support, advocacy, translation, teaching, or representation it involved. Conversely, they may list a service organization but overlook the planning, research, or communication work they performed.
The paired evidence showed that later inventories often made functions visible even when the underlying organization was already known. This was usually reframing rather than discovery, but it points to the same practical lesson: the activity name is not the activity.
6.5 A missing line is not automatically a valuable line
The study intentionally does not rank the seven verified entries. Discovering a pre-existing experience does not mean it should displace another item. Application space is limited, and relevance depends on duration, meaning, contribution, context, and the rest of the student’s application. The appropriate decision may be to include the experience, combine it with another, mention it elsewhere, or leave it out.
This distinction protects against a common error in advising discourse: equating more content with better content. Common App’s own worksheet says students may list up to ten activities but do not need to enter ten. The purpose is to communicate meaningful pursuits, interests, and passions, not to fill every slot.
Why Most Apparent Additions Were Not Omissions
7.1 Students began new work
The largest non-matched category was genuinely new activity. This is unsurprising given the observation period. Students entered new grade levels and summer periods. They gained access to programs, pursued research, started jobs, and developed projects. A baseline should be treated as a snapshot, not a permanent profile.
The finding has implications for counseling schedules. A one-time inventory early in the process will become stale. Families need a light update routine: note what began, ended, expanded, paused, or changed. That update is not an invitation to manufacture new commitments for application purposes. It is ordinary documentation of a changing life.
7.2 Roles evolved within existing commitments
Among the 54 matched entries, many later descriptions showed progression. A student could move from participant to leader, from learner to instructor, or from a general team role to a specialized responsibility. Those developments may justify updated language, but the continuity of the underlying activity should remain legible.
Progression is often more informative than title accumulation. A strong inventory records what changed: whom the student supported, what system was built, what problem was addressed, what output was completed, or what responsibility became independent. Titles without functions can obscure rather than clarify.
7.3 Related experiences were merged or separated
Application constraints require architecture. Two closely related roles may be combined to avoid redundancy. One broad line may be separated when it contains distinct commitments with different dates and contributions. Neither operation creates experience; it changes the map.
For research, this means entries are not natural units. The same life history can yield eight, ten, or twelve lines depending on editorial choices. Counting lines without matching underlying experience can confuse formatting with substance.
7.4 Honors and interests were sometimes reframed
Earlier accounts sometimes placed an experience under honors, education, skills, or interests rather than activities. A later inventory could elevate that same material into an activity entry. The study treated it as represented when the underlying experience was already visible. This conservative rule avoids crediting the later process for discovering something that was present under another heading.
The distinction is useful for families. A résumé audit should scan every section, not only “Activities.” An award may point to sustained participation. An interest line may contain a meaningful independent practice. A skill may have been developed through a project worth describing. The task is to trace the underlying work while avoiding duplicate claims.
7.5 Some timelines remained unresolved
Indeterminate items are not analytical failures; they reveal the limits of retrospective evidence. If a later entry lacks a start date and the earlier account is silent, no amount of confident prose can establish onset. Families can reduce this problem by preserving basic dates, but researchers should not erase it after the fact.
The ten indeterminate entries also caution against precise-sounding anecdotes. It would be easy to describe a hobby as something “the student had done for years” or a club as “long overlooked.” Unless the source establishes that fact, it should not enter the paper.
A Practical Activity-Reconstruction Method
The findings support a structured process that separates memory, verification, and selection. The following method is offered as practical advice, not as an experimentally validated intervention.
8.1 Pass one: reconstruct time, not prestige
Begin with a calendar or grade-by-grade timeline. For each school year and summer, ask where the student regularly spent time outside class. Do not begin by asking for the “best” activities. Prestige filtering will narrow recall too early.
Useful time anchors include the start and end of each school year, summers, moves, school changes, family transitions, jobs, competition seasons, religious calendars, major projects, and recurring community events. The goal is to generate raw commitments before judging them.
8.2 Pass two: scan life contexts
Use contexts rather than labels. Ask about school, home, work, neighborhood, online communities, faith or culture, athletics, arts, independent learning, research, service, and summers. A context matrix reduces institution bias because it does not require the student to decide in advance what counts as extracurricular.
At school, ask about organizations, teams, publications, student government, peer support, performances, competitions, labs, and roles that occurred inside a class but extended beyond ordinary coursework. At home, ask about caregiving, translation, transportation, household management, family business, and recurring responsibilities. At work, include paid employment, informal jobs, seasonal work, and sustained entrepreneurial activity. Online, ask about creation, moderation, collaboration, teaching, open-source work, communities, and audiences—not passive consumption alone.
8.3 Pass three: convert nouns into verbs
For every candidate experience, list what the student actually did. “Robotics” is a noun. Designing, fabricating, documenting, testing, scouting, fundraising, teaching, and troubleshooting are actions. “Volunteer” is a status. Delivering, translating, tutoring, organizing, researching, or maintaining are actions.
Verbs help reveal whether two entries are duplicates and whether a seemingly informal experience involved meaningful responsibility. They also make unsupported inflation easier to detect. If the description contains ambitious outcomes but no concrete action, it may be too abstract.
8.4 Pass four: verify chronology
Record start month or season, end month or “ongoing,” grade levels, and major role changes. If exact dates are unavailable, use honest approximations and mark uncertainty. Do not backfill precision merely because an application field asks for it.
Chronology serves three purposes. It prevents new experiences from being described as long-standing, helps estimate weeks and hours realistically, and allows advisers or researchers to distinguish later development from earlier omission. The present study’s fourfold difference between apparent and verified omissions demonstrates why this step matters.
8.5 Pass five: verify scope and evidence
For each statement, ask what supports it. Counts may be drawn from attendance, published outputs, schedules, logs, fundraising records, team rosters, or other contemporaneous evidence. If a number is an estimate, label it internally and use a reasonable rounded figure. If no support exists, describe the contribution without a number.
Avoid invented impact metrics, inflated global claims, and causal statements the student cannot substantiate. “Created a resource used by peers” is stronger and safer than an unsupported claim that the resource transformed outcomes. Accuracy is not merely an ethical constraint; it produces clearer writing.
8.6 Pass six: identify overlap
Create an underlying-experience key. If two candidate entries refer to the same organization, project, and period, decide whether they are truly distinct. Separate entries should usually have different functions, time commitments, or outputs. Otherwise combine them.
Also scan honors, skills, interests, employment, research, and education for hidden duplicates. A competition award can support an activity description without becoming a second activity. A summer program may belong under education or activities depending on the application, but it should not be counted twice in an analytic inventory.
8.7 Pass seven: choose, do not fill
Only after the full inventory is verified should the student select entries for a particular application. Consider meaning to the student, duration, responsibility, contribution, context, connection to other parts of the application, and redundancy. There is no requirement to use all available spaces simply because they exist.
Selection should preserve a truthful picture, not manufacture a theme. A coherent application can include variety. A student interested in engineering may still include paid work, family care, art, or community activity when those commitments are meaningful.
8.8 Pass eight: write at the right level
A concise activity description generally needs four elements: role, action, scope, and result or learning where support exists. Not every line needs all four. The priority is to make the student’s direct involvement understandable.
Consider this synthetic example: “Community garden volunteer” becomes “Planned weekly harvests with six volunteers; translated planting guides; delivered produce to a neighborhood pantry.” The revision is not stronger because it sounds grander. It is stronger because it replaces a vague label with verifiable actions.
Another synthetic example: “Helped at family store” becomes “Closed register three evenings weekly, translated for customers, updated inventory, and trained a younger sibling.” The activity was always there. The inventory process recognized its functions.
8.9 Pass nine: conduct a final integrity review
Before submission, the student should be able to explain every title, date, hour estimate, number, and outcome in ordinary language. Check that future work is not written as completed, team results are not claimed as individual results, and organizational scale is not confused with the student’s reach.
The student should own the final wording. Advising can ask questions, help organize evidence, and improve clarity, but it should not replace the student’s knowledge or voice.
Implications For Families, Schools, And Advisers
9.1 For families
Families can help by supplying memory anchors without turning the process into a competition. A parent or guardian may remember transportation routines, work schedules, cultural or religious commitments, family responsibilities, and long-running hobbies that the student normalizes. The student should still decide what is accurate and meaningful.
The most useful family question is often “Where did your time go?” rather than “What impressive activity are you missing?” The first invites reconstruction. The second invites performance.
Families should also distinguish encouragement from authorship. They can help locate dates or evidence, but the application should reflect the student’s experience and language. A polished line that the student cannot explain is a warning sign.
9.2 For school counselors and teachers
Schools can reduce reporting inequity by broadening the examples used in application workshops. If every sample features elected leadership, selective research, and national awards, students may infer that paid work, caregiving, hobbies, local service, and independent learning do not count. Official Common App guidance explicitly includes a broader set.
A grade-by-grade activity log can also help. It need not be elaborate. A simple annual exercise recording commitments, dates, responsibilities, and major outputs would reduce senior-year reconstruction and improve recommendation context.
Counselors should be cautious about quantity benchmarks. Common App’s large-scale research shows that reported counts differ across demographic and socioeconomic groups and that opportunity structures matter. Advising students to “fill all ten” may amplify those differences and encourage thin or duplicative entries.
9.3 For independent advisers
Advisers should document their matching logic when claiming that a process uncovered activities. The present study suggests three separate metrics: matched/reframed entries, genuinely new entries, and verified previously unlisted entries. Combining them creates an inflated impact measure.
Advisers should also avoid using an activity-discovery process as evidence of improved admission chances. The activity section is one component of holistic review, and NACAC’s Fall 2023 survey shows wide variation in how colleges rate extracurricular activities and work. In that survey, 6.5% of responding four-year colleges assigned extracurricular activities “considerable importance,” 44.3% “moderate importance,” 30.8% “limited importance,” and 18.4% “no importance.” Work was rated considerable by 2.2%, moderate by 30.8%, limited by 40.0%, and not important by 27.0%. Those national institutional ratings do not predict any individual decision, but they caution against sweeping causal claims.
9.4 For application platforms
Platform design shapes recall. Common App’s recent emphasis on activities and experiences, responsibilities, and circumstances recognizes that students’ commitments extend beyond conventional clubs. Clear examples, flexible categories, and prompts about family and work can improve representational breadth.
However, more fields do not automatically solve the classification problem. Students still need to translate their lives into the platform’s vocabulary. Guidance should reinforce that examples are invitations, not rankings, and that applicants need not maximize quantity.
9.5 For researchers
Future studies should treat the activity list as longitudinal, nested data. Entries are nested within students; drafts are nested within entries; and opportunities unfold over time. A robust design would prospectively collect baseline inventories, onset dates, adviser prompts, final submissions, and independent coding. It would distinguish discovery from description quality and would not use admissions outcomes as a simple validation label.
Researchers should also measure inter-rater reliability. The current study used a documented conservative protocol but did not include a blinded second coder. A larger study could pre-register definitions, double-code a substantial subset, report agreement, and resolve disagreements without access to outcomes.
Comparison With External Research
10.1 Common App’s definition supports broad reconstruction
Common App’s official materials are consistent with the study’s practical premise that relevant experiences extend beyond school clubs. Its first-year guidance names work, hobbies, clubs, community engagement, responsibilities, and circumstances. Its activities worksheet specifically asks students to consider family responsibilities, jobs, internships, volunteering, hobbies, and sports. It also says applicants may list up to ten activities but need not enter ten.
That guidance does not verify the internal finding that seven entries were previously unlisted. It does, however, support the content validity of a broad discovery checklist. The categories recommended here are not inventions designed to increase list length; they reflect the platform’s own framing of experiences outside the classroom.
10.2 National reporting disparities complicate quantity comparisons
Common App’s 2023 research summary examined approximately 860,000 applicants and nearly six million activity descriptions from the 2018–19 and 2019–20 seasons. It found substantial differences in the number of activities reported. Continuing-generation applicants reported 7.45 activities on average compared with 5.44 among first-generation applicants; applicants who did not receive fee waivers reported 7.46 compared with 5.51 among recipients; and private-school students reported 7.87 compared with 6.71 among public-school students.
The authors emphasized that financial, cultural, logistical, school, and college-preparation factors contribute to these gaps. They suggested careful contextualization and greater focus on quality rather than quantity. This evidence argues against using the number of discovered entries as a performance metric for students or advisers.
The Ivy Institute sample cannot be directly compared with those national averages. The internal total is a working-inventory count among seven selected advising cases, not the number finally reported by a national applicant cohort. Different populations, years, definitions, and stages make direct benchmarking invalid.
10.3 NACAC’s application guidance recognizes multiple experience types
NACAC’s overview of application components lists extracurricular, personal, and volunteer activities separately from employment, internships, and summer activities. That separation reinforces the need to ask across contexts. A student who hears only “extracurriculars” may not automatically surface employment or summer work.
NACAC’s admission-factor survey also shows that colleges vary in the importance assigned to activities and work. This variation supports a restrained interpretation: documenting an experience accurately can improve completeness, but no activity type guarantees an outcome.
10.4 External research offers context, not validation
None of the external sources reviewed examined The Ivy Institute’s seven paired cases. They cannot confirm the internal counts, prove the advising process caused an addition, or establish that the resulting lists were preferred by admission officers. Their role is narrower: they show that broad activity categories are officially recognized, that reporting quantity varies systematically across groups, and that colleges differ in the weight assigned to activities.
The distinction between contextual support and independent validation is essential. A finding becomes more plausible when it aligns with a broader literature, but it does not become independently replicated until another study examines comparable paired evidence using a similar method.
10.5 What the external evidence changes—and what it does not
The external sources affect interpretation in three specific ways. First, they broaden the domain of relevant experience. When the application platform itself names work, responsibilities, hobbies, volunteering, and community engagement, a discovery process that asks only about clubs is incomplete by the platform’s own standard. Second, the national research shows that reported activity count reflects opportunity and preparation as well as individual participation. This undermines any simple assumption that more entries mean more merit. Third, the admission-factor survey shows that institutional weighting varies, so the same activity inventory may be read differently across colleges.
None of those points determines which of the seven internal entries should have been submitted. They do not prove that a student was disadvantaged by an earlier omission, and they do not identify an ideal number of activities. External evidence can guide the questions asked; it cannot substitute for the student’s context or for college-specific instructions.
10.6 Sensitivity checks
The study conducted several conceptual sensitivity checks to test whether the headline depended on an aggressive rule.
The first check used the broadest possible definition of apparent addition: every later entry without a baseline counterpart. That produced 30 additions, or 35.7% of the later inventory. This is the upper-bound surface comparison. It is transparent but substantively misleading if labeled “forgotten,” because it includes new and undated experiences.
The second check removed the thirteen items known to have begun later but treated all ten indeterminate items as if they had existed earlier. That would produce a maximum of 17 potentially pre-existing omissions, or 20.2% of all later entries. This is not the preferred estimate because it assumes what the chronology does not show. It is useful as an uncertainty bound: even under the most omission-favorable treatment of missing dates, the count remains far below thirty.
The third check applied the strict rule used in the report: count only entries with affirmative evidence of pre-baseline onset. That produced seven, or 8.3% of all later entries. The strict estimate sacrifices recall to protect precision. Some of the ten indeterminate entries may truly have been omitted earlier, but the study refuses to claim them without support.
The plausible observed range is therefore seven to seventeen depending on how uncertainty is handled, with seven as the verified count. This range does not estimate a population parameter. It describes the effect of alternative assumptions within the seven cases. Publishing both the strict result and the uncertainty bound helps readers see that the conclusion is not a single magic number.
The fourth check concerned unit definition. If duplicate drafts and optional alternatives had been counted, students with extensive revision histories would contribute many more entries. Excluding those versions avoids making documentation intensity look like activity growth. If a later inventory contained multiple descriptions of the same underlying commitment, the study counted one activity.
The fifth check concerned matching. The protocol matched by underlying experience rather than exact wording. An exact-text rule would move many of the 54 represented entries into the apparent-addition category, dramatically inflating change. The underlying-experience rule is more labor-intensive but more faithful to the research question.
10.7 Reproducible arithmetic
The public calculations require only four counts. The later inventory denominator is 84. The represented-or-reframed count is 54, calculated as 54 divided by 84, or 64.3%. The genuinely new count is 13, or 15.5%. The indeterminate count is 10, or 11.9%. The verified previously unlisted count is 7, or 8.3%. The four categories sum to 84 and their rounded percentages sum to 100.0%.
The apparent-addition denominator is the sum of the three non-matched chronology categories: 13 plus 10 plus 7 equals 30. Within those 30, genuinely new entries account for 43.3%, indeterminate entries for 33.3%, and verified previously unlisted entries for 23.3%. Thirty divided by seven equals approximately 4.29, which is the basis for saying that the surface method would overstate the verified count by more than fourfold.
No confidence interval is reported. The seven cases were not randomly sampled from a defined population, and entries are clustered within students. A conventional binomial interval around 7 of 84 would create a false impression that 84 independent activities were sampled. They were not. Descriptive transparency is more appropriate than inferential decoration.
10.8 A decision matrix for families
Once a family has reconstructed a broad inventory, each candidate experience can be placed in a simple decision matrix.
The first dimension is evidence. Is the activity’s existence, timing, and scope known well enough for accurate reporting? A meaningful experience with uncertain hours can still be reported using a reasonable estimate, but a claimed award, audience, outcome, or leadership role should be verifiable.
The second dimension is distinctiveness within the student’s own application. Distinctiveness does not mean rarity in the national applicant pool. It means that the entry contributes information not already communicated by another line. Two entries may both involve science, for example, but one may show long-term team leadership and the other independent investigation. Conversely, three titles within the same organization may be better combined.
The third dimension is significance to the student. Significance can come from time, responsibility, growth, contribution, identity, intellectual engagement, or family context. It need not come from selectivity or prestige. A paid job may be central because it supported a household or developed responsibility. A hobby may be central because it sustained years of creative production.
The fourth dimension is fit with the application’s instructions. Some platforms provide separate spaces for employment, family responsibilities, portfolios, research, or additional information. A complete personal inventory can be routed differently across applications. Families should follow each institution’s directions rather than copy one architecture everywhere.
Candidate experiences with strong evidence, distinct information, personal significance, and a clear application home are natural priorities. Experiences with weak evidence should be corrected or removed. Experiences that duplicate another entry should be combined or deprioritized. Experiences that are meaningful but private require the student’s consent and judgment; completeness never overrides autonomy.
10.9 The difference between discovery and embellishment
A broad inventory method can be misused. Because the application rewards concise descriptions, there is pressure to package ordinary actions with inflated titles and dramatic metrics. That is not the purpose of discovery. Discovery recognizes work that occurred. Embellishment increases the apparent scale, status, or impact beyond what occurred.
Several guardrails follow. Position titles should reflect actual roles, not invented executive language. Team achievements should be attributed to the team unless the student’s contribution can be separated. Organization-wide reach should not be described as the student’s personal reach. Planned work should not be written in past tense. Estimates should be reasonable and internally consistent. Superlatives such as “first,” “largest,” or “global” require evidence and relevance.
The best test is the explanation test: could the student describe the activity naturally in a conversation, explain where the figures came from, and distinguish personal action from group context? If not, the line needs revision.
10.10 A model annual inventory routine
The research suggests that families do not need a high-pressure senior-year excavation if they maintain a modest annual routine. At the end of each school year, a student can record commitments, approximate dates, hours during a typical week, major outputs, role changes, and one sentence about what mattered. The record should include work, family responsibilities, independent projects, hobbies, and community participation alongside school activities.
At the end of summer, the student can add new programs, jobs, travel with substantive responsibilities, independent learning, and projects. At the beginning of the next year, planned activities should be marked as planned rather than completed. A brief winter update can record leadership transitions or completed outputs.
This routine has benefits beyond college applications. It can support recommendation requests, scholarship forms, employment applications, and personal reflection. It also reduces the risk of reconstructing hours from memory years later. However, the routine should remain a record, not a performance dashboard. Students should not feel obligated to maximize every field.
10.11 Future research agenda
A stronger follow-up study would recruit a larger consecutive cohort and establish a common baseline instrument before advising begins. Students would complete a broad inventory, record confidence in each date and metric, and identify which experiences they initially considered “activities.” Advisers would then use a standardized prompt sequence. New entries would be coded prospectively as recalled, newly begun, split, merged, or re-described.
Independent reviewers, blinded to student outcomes and whether an entry appeared before or after advising, could code the distinctness, evidentiary support, and clarity of descriptions. Inter-rater agreement would be reported. Final submissions could be verified separately from working inventories. Students could also be asked why an experience was absent earlier, allowing the research to distinguish forgetting from perceived irrelevance, privacy, purpose mismatch, or space constraints.
An equity-focused extension could examine whether broad prompts surface different categories across socioeconomic background, first-generation status, school resources, paid-work intensity, or family responsibilities. Such work would require careful consent, sufficient sample sizes, and suppression rules. It should avoid treating historically underreported responsibilities as deficiencies in the student.
A process evaluation could test whether annual documentation reduces indeterminate chronology and senior-year stress. A causal study of admission outcomes would be substantially harder. Random assignment to a list-building process, adequate sample size, college-level outcome modeling, and controls for selection would be necessary, and even then the activity list could not be isolated cleanly from the rest of the application. The present report does not make that leap.
Limitations And Research Ethics
11.1 Small, selected sample
Seven students are too few for population estimates or stable subgroup comparisons. The cases were selected because pairable materials existed, not through random sampling. Students with less documentation, shorter advising relationships, or simpler records may be underrepresented. The findings may describe intensive application preparation more than ordinary application behavior.
11.2 Variable baselines
The earlier accounts differed in purpose and completeness. Some may have been designed as broad résumés; others may have emphasized selected achievements. The study’s “previously unlisted” category means absent from that observed baseline, not absent from every prior self-description.
11.3 Chronology gaps
Ten later-only entries could not be classified by onset. More complete dates might move some into previously unlisted or genuinely new. The published counts should therefore be read as observed categories under a conservative rule, not as the unknowable true distribution.
11.4 No independent second coder
The study used explicit rules and retained a restricted item-level ledger, but it did not employ a blinded independent coder. Matching underlying experiences sometimes requires judgment. Future replication should measure coder agreement.
11.5 No final-submission verification for every entry
The study analyzes principal later inventories, not verified submissions for every student and college. Some entries may have been considered and then excluded. Accordingly, the paper uses “later inventory” rather than “final application” except when discussing general application mechanics.
11.6 No causal design
There was no control group, random assignment, or counterfactual. Students changed over time, and many activities began after baseline. The research cannot determine how many entries would have surfaced without advising or whether the process improved an admission outcome.
11.7 No quality or value score
Entries were counted as experiences, not judged for strength. A sustained hobby and a formal internship each count as one item in this analysis, even though duration, meaning, context, and application relevance differ. The count measures reporting transition only.
11.8 Privacy limits detail
Protecting minors and families required the removal of names, quotations, exact organizations, and student-level combinations. That reduces narrative richness and prevents readers from auditing the raw evidence. The restricted ledger preserves traceability for authorized review, while the public appendix preserves reproducible aggregate arithmetic.
11.9 Company interest
The Ivy Institute conducted, funded, and is publishing this study. It has a commercial interest in advising services and in research that portrays structured reflection as useful. The conservative chronology rule, disclosure, public calculations, and limitation language reduce but do not eliminate this conflict. Readers should prefer independent replication before generalizing.
11.10 Ethical interpretation
The findings should not be used to pressure students into creating more activities, rebranding normal life as extraordinary, or disclosing private family circumstances they do not wish to share. Broad prompts should expand recognition, not compel exposure. Students should decide what is relevant and appropriate within each application’s instructions.
Conclusion
The paired records show both the promise and the danger of before-and-after activity research. Later inventories were meaningfully different, but the differences did not have one cause. Most entries were continuations of experiences already present. Among the thirty apparent additions, thirteen were genuinely new and ten could not be dated confidently. Seven were demonstrably pre-existing but previously unlisted.
That seven-entry result is useful, but the more important result is the discipline required to reach it. Without matching and chronology, the apparent omission count would have been more than four times larger. Research on advising should not treat every new line as evidence that counseling discovered hidden value.
For families, the practical message is constructive. Students often need a wider vocabulary for their own time. Independent creation, work, community contribution, research, student representation, household responsibility, and sustained hobbies can be meaningful parts of an application. A structured, grade-by-grade reconstruction can surface them. But discovery must be followed by verification and selection. The goal is an accurate account, not the longest possible list.
For advisers and researchers, the message is stricter. Separate experience from wording, development from discovery, and uncertainty from absence. Report denominators. Preserve dates. Do not equate a refined list with admission impact. And when the evidence cannot tell whether an experience existed earlier, say so.
In that sense, the activities students “forget” are only part of the story. The fuller story is how adolescent lives resist being compressed into application fields—and how careful methods can help describe those lives without overstating what the record proves.
References
Common App. “Application guide for first-year students.” Current page accessed November 9, 2026. https://www.commonapp.org/apply/first-year-students/
Common App. “Approaching the Activities section.” Common App Ready, 2023; toolkit revision noted October 15, 2026. https://www.commonapp.org/files/FY-Apply-College-Activities-Section-Common-App.pdf
Common App. “Common App user interface updates 2025–2026.” https://www.commonapp.org/files/Common-App-UI-updates.pdf
Common App Data Analytics and Research. “Trends and disparities in extracurricular activity reporting.” April 18, 2023. https://www.commonapp.org/files/common-app-report-trends-disparities-extracurricular-activities.pdf
Common App. “What’s new with Common App: 2025–26 updates.” May 22, 2025. https://www.commonapp.org/files/Whats-New-25-26.pdf
National Association for College Admission Counseling. “Factors in the Admission Decision.” Fall 2023 survey results, current page accessed November 9, 2026. https://www.nacacnet.org/factors-in-the-admission-decision/
National Association for College Admission Counseling. “Parts of a College Application.” Current page accessed November 9, 2026. https://www.nacacnet.org/parts-of-a-college-application/
The Ivy Institute, internal materials, data, and information.
