What Not To Do
12. What Not To Do
Zara Okonkwo, with deadlines approaching, the biggest risks in your application are not about grades or test scores. Your 3.94 GPA and 1530 SAT already show academic readiness. The real danger is how your work in data science or statistics is interpreted by readers who only spend a few minutes with your file. Admissions officers will look for clear evidence that you personally understand and apply technical methods. When that evidence is missing or vague, strong work can be discounted quickly.
The committee flagged several patterns that frequently weaken otherwise competitive data‑focused applications. Avoid the following pitfalls carefully as you finalize your materials.
Do Not Let a Civic Impact Story Replace Technical Evidence
Applications centered on social impact—especially data used for public good—can be compelling. However, one of the risks in this area is presenting a project primarily as a community or civic story without showing the statistical or computational work that produced the results.
If an admissions reader finishes your description knowing why the issue matters but not how the analysis was performed, they may assume the technical component was minimal. This is particularly risky for data science applicants to highly technical programs such as UC Berkeley, Carnegie Mellon, and Georgia Tech.
Common mistakes that weaken an application:
- Describing the social problem in detail but mentioning the analysis only briefly.
- Focusing on outcomes (“we helped the community understand X”) instead of the statistical process used.
- Leaving out the models, algorithms, or analytical methods used.
The result is that a potentially strong project reads like advocacy rather than technical investigation. At highly selective technical programs, that distinction matters.
If you include civic‑oriented data work, avoid letting the narrative become purely about impact. The admissions reader must clearly see the analytical depth behind the results.
Do Not Leave Open‑Source Work Vague
If you plan to reference open‑source code, shared datasets, or collaborative repositories, vague descriptions can unintentionally weaken your profile.
Admissions reviewers often encounter applicants who claim “open‑source work” but provide almost no context. When the scope is unclear, readers frequently assume the contribution was small.
Risky phrasing to avoid:
- “Contributed to an open‑source data project.”
- “Worked on a GitHub repository with others.”
- “Helped develop a data analysis tool.”
Without context, these statements raise more questions than they answer. Reviewers may wonder:
- Was this a major contribution or a small fix?
- Did you design the statistical approach or simply implement instructions?
- Was the code actually used by others?
If open‑source work is included anywhere in your application materials and remains vaguely described, readers often default to assuming minimal impact. That is a preventable interpretation problem.
If you do reference this type of work, clarity about scope and ownership becomes essential. Otherwise it can quietly undermine what might otherwise be a strong technical signal.
Do Not Assume Team Competitions Prove Individual Skill
Team competitions—especially in data science, math, or analytics—are common activities among applicants interested in statistics or machine learning. However, admissions readers do not automatically interpret team results as evidence of individual technical ability.
If a competition result is presented without explanation, the reader has no way to determine what you personally did.
Problematic patterns include:
- Listing a team placement with no explanation of your role.
- Describing the competition outcome but not your contribution.
- Using “we built” or “our team analyzed” without clarifying your responsibilities.
From the reviewer’s perspective, several possibilities exist:
- You may have led the modeling work.
- You may have handled data cleaning or visualization.
- You may have had a non‑technical role.
If the application never clarifies this, readers cannot confidently attribute the technical achievement to you. For applicants to programs like CMU or Berkeley, that ambiguity can weaken the perceived strength of the activity.
Team work is valuable, but the admissions reader must understand your individual contribution.
Do Not Submit Data Projects That Look Like Simple Visualization
Another common mistake among data science applicants is presenting projects that appear to focus primarily on aggregation or visualization.
Dashboards, charts, and visual summaries are useful tools. But if the project description emphasizes visual output without mentioning deeper statistical analysis, admissions readers may interpret the project as introductory rather than advanced.
Signals that trigger this interpretation include:
- Descriptions focused on charts, dashboards, or plots.
- Mentioning data collection and visualization but not modeling or inference.
- No explanation of statistical reasoning behind the results.
This can unintentionally place a project in the category of “basic data exploration” rather than serious analytical work.
For highly quantitative programs, readers typically look for evidence of:
- statistical reasoning
- modeling approaches
- methodological choices
- interpretation of results
If those elements are absent from the description, the project can appear superficial even if substantial work actually occurred.
Do Not Leave Technical Work Implicit
A broader theme connects all of these pitfalls: assuming the reader will infer technical depth without seeing it explicitly described.
Admissions officers reading thousands of applications rarely infer complexity that is not directly visible. When technical work is implied rather than clearly stated, the safest interpretation from their perspective is that the analysis was limited.
This is especially important for applicants targeting quantitative majors. Programs in data science and statistics expect clear evidence that applicants understand analytical tools and methods.
If the application narrative centers on outcomes, collaboration, or impact but leaves the analytical process unclear, the file may read as less technically rigorous than intended.
Application Execution Calendar (Avoidance Focus)
| Month | Actions to Prevent These Pitfalls |
|---|---|
| September |
|
| October |
|
| November |
|
If you avoid these specific traps, your technical interests will come across much more clearly to readers at Berkeley, Carnegie Mellon, and Georgia Tech. The goal is simple but critical: make sure every reviewer can immediately see the analytical thinking behind your work.