What Reviewers Look For in the PhD SOP — Research Dossier Read
This rubric evaluates whether your statement uses your past research to show how you think and what you can do, then connects that record to a problem you want to explore next. It also looks for a clear explanation of why that problem matters, why it remains difficult or unresolved, and what resources at the department would help you pursue it. Strong writing is precise, evidence-based, and focused on your own role in the work.
What We Check
Where the statement first names research content and how much of the document goes to material that is neither the research record, the forward problem, nor what the work would need from the department.
Whether the research episodes the statement describes carry a complete account: the problem the work addressed, something the applicant tried or built, a decision about how the work was done, and what came of it.
How much of the described work is the applicant's own - which sentences separate what they did from what the group did, and whether anything in the record began with them.
Whether the statement names one problem it wants to take on next and characterizes it: what the question is, why it matters, why it is hard, and why it is still open.
Whether the record the statement describes reaches the problem it wants to work on next - whether a reader can point to the sentence saying what in the applicant's own work makes them able to take this on.
Whether the statement says what the forward problem needs and names something at the receiving department that meets it.
Mistakes We Flag
- Opening with childhood interest or a class epiphany
- An early interest, film, book, or favorite class can delay the research focus. If you include it, connect it directly to a research problem.
- Listing grades, awards, jobs, or lab names
- A résumé-style list does not show what you investigated, tried, decided, or learned. Give at least one piece of work enough detail to make your contribution clear.
- Naming techniques without the research question
- Methods, software, and instruments need context. Explain the question they helped address and what the work produced or revealed.
- Describing only the team’s work
- Repeatedly saying “we” can leave your contribution unclear. State what you personally built, decided, changed, or did in response to a challenge.
- Giving a broad future topic instead of a problem
- A subfield or application area alone does not explain what you want to investigate. Name the specific unanswered question, what is already known, and where that knowledge stops.
How this rubric reads your essay
- Built from published reader guidance — the sources are listed below.
- Scored per section against defined criteria, not general impressions.
- Calibrated for consistent scoring, so a better draft shows up in the score.
References
VERIFIED 2026-09-03 by direct fetch (desktop-UA curl, HTTP 200) with an exact-substring match on the quoted words after tag-stripping and whitespace normalization; no proxy, no cache. The primary evaluator sheet this rubric mirrors: background "can provide useful context, but keep it relatively short; spend most of your time on the research you've done and what you hope to do in the future"; the prior-work account ("describe the problem you confronted, argue that it is important and was challenging to solve, compare (briefly) to prior work, and explain your technical solution and how you validated it"); the forward problem ("explain what is the scientific question, why it is important, why it is hard, and why it is open" - with an approach optional: "You might give some ideas for an approach to solving it, though characterizing the problem well is often critical"); and naming as conditional ("Mention the names of specific faculty if you know you want to work with them").
VERIFIED 2026-09-03 by direct PDF fetch (desktop-UA curl, HTTP 200) plus pdftotext extraction, with an exact-substring match on the quoted words; no proxy, no cache. The document header reads "Last updated 2014". Supports the genre correction ("what admission committees are really looking for is a research statement"), the episode account ("What was the problem you were trying to solve? Why was it important? What approaches did you try? What did you learn? It's fine to say that you were unable to fully solve your problem."), the non-commitment of the stated direction ("the research statement is not a commitment to do research in that area. A third of all applicants end up working in an area different from that which they described"), the access floor ("prior research experience does not mean that you need to have published a paper. It does not even mean that your research needs to have yielded a result"), and the named failure modes "The grade regurgitator" and "The boy genius".
VERIFIED 2026-09-03 by direct fetch (desktop-UA curl, HTTP 200) with an exact-substring match on the quoted words after tag-stripping and whitespace normalization; no proxy, no cache. Supports the space rule and, on the same page, the protections that keep it from becoming a disclosure penalty: "Apart from the above, an overly-lengthy personal background is generally not helpful"; "We recognize that not all students were able to access such opportunities. We welcome other work or personal experiences that, even if not directly relevant, demonstrate your experience in overcoming challenges or working with others"; "You may also choose to explain any unusual circumstances and how they have affected you, such as gaps in your resume or challenges you have faced"; goals "may be kept brief if included"; and "Between one and two pages is generally sufficient".
VERIFIED 2026-09-03 by direct fetch (desktop-UA curl, HTTP 200) with an exact-substring match on the quoted words after tag-stripping and whitespace normalization; no proxy, no cache. Sixteen named MIT faculty on one page (count corrected 2026-09-04 by the Lane C source-verification pass; the earlier note said eleven). Supports the ownership axis (Adam Chlipala: "ideally applicants can point to cases where they were the ones driving that brainstorming, not just implementing ideas coming from supervisors"), the non-binding direction (Chlipala: "don't worry that anyone will hold you to working on any of the specific ideas you list!"), the capability-evidence reading (Piotr Indyk: "the most important aspect of the application is the evidence of research skills"), the access carve-out (Indyk: "it is understandable that not all applicants have the same opportunities to pursue undergraduate research"; Karl Berggren: "This is the place where I want to know if someone has faced exceptional challenges"), and the unusual-route credit (Leslie Kaelbling). The page also carries the dissent this rubric declines (Polina Golland: keep past projects "to the minimum").
VERIFIED 2026-09-03 by direct fetch (desktop-UA curl, HTTP 200) with an exact-substring match on the quoted words after tag-stripping and whitespace normalization; no proxy, no cache. Supports the per-episode depth bar ("One should go into a little depth with your research experience — saying exactly what you did") and carries a worked example that separates the applicant's own part from a colleague's and states a checkable outcome. Also the reason this rubric carries no cross-document-redundancy flag: "there is nothing wrong with repetition. The statement of purpose is the most visible place to put things which you want seen."
VERIFIED 2026-09-03 by direct fetch (desktop-UA curl, HTTP 200) with an exact-substring match on the quoted words after tag-stripping and whitespace normalization; no proxy, no cache. Supports scoring a provisional forward problem at full strength and carrying no feasibility or timeline axis: "Its [sic] not necessary to have a specific thesis proposal, and odds are if you try to pretend you have one when you really don't you'll come off as sounding naive" (the missing apostrophe is the page's own). The page carries the author's disclaimer that these are opinions.
VERIFIED 2026-09-03 by direct fetch (desktop-UA curl, HTTP 200) with an exact-substring match on the quoted words after tag-stripping and whitespace normalization; no proxy, no cache. The published opposite pole to adviser-naming, and the reason this rubric scores an unnamed departmental route and a named person on the same terms: "we admit students to the statistics department to work in statistics, but not to work with a particular faculty member."
VERIFIED 2026-09-03 by direct PDF fetch (desktop-UA curl, HTTP 200) plus pdftotext extraction, with an exact-substring match on the quoted words; no proxy, no cache. Supports the technique-list red flag ("Don't just list techniques, needs to talk about techniques within research, talk about the science"), the genre stance ("PhD applications are not narratives, not telling your story, its about research"), and the single-person caution ("Don't ever apply to a program for only one professor"). NOTE: this page's "(90%)" figure is scoped on the page to obstacle material specifically and is deliberately NOT used as an allocation anchor here.
VERIFIED 2026-09-03 by direct fetch (desktop-UA curl, HTTP 200) with an exact-substring match on the quoted words after tag-stripping and whitespace normalization; no proxy, no cache. Supports the genre name and the CV-recap red flag: "You are asked to provide a Statement of Purpose, which we expect to be a Research Statement"; "The Research Statement should provide context for your CV"; "This statement should be around 2 single-spaced pages in length." The same page is a naming-and-contact-expecting department, which is why naming is a fully credited route here.
VERIFIED 2026-09-03 by direct fetch (desktop-UA curl, HTTP 200) with an exact-substring match on the quoted words after tag-stripping and whitespace normalization; no proxy, no cache. Non-CS confirmation that the prior-work account is the primary payload: "the admissions committees are interested in obtaining as much information as possible about your previous research experiences."
VERIFIED 2026-09-03 by direct fetch (desktop-UA curl, HTTP 200) with an exact-substring match on the quoted words after tag-stripping and whitespace normalization; no proxy, no cache. Supports the professed-passion red flag outside the lab sciences: "don't spend too much time describing your passion for economics. This will be unlikely to persuade admissions committees"; "Get to specifics as soon as possible."
VERIFIED 2026-09-03 by direct fetch (desktop-UA curl, HTTP 200) with an exact-substring match on the quoted words after tag-stripping and whitespace normalization; no proxy, no cache. The named-opener catalog: "Professors don't care that you got the first computer on your street and started messing with it when you were really young" (quoted here as the full sentence, per the corpus correction of 2026-08-20), and the coursework-epiphany opener ("If your essay starts off with talking about how you took a class that you really liked and it made you want to get a Ph.D., then you're doing it wrong").
VERIFIED 2026-09-03 by direct fetch (desktop-UA curl, HTTP 200) with an exact-substring match on the quoted words after tag-stripping and whitespace normalization; no proxy, no cache. A humanities warrant for the capability-evidence spine ("First and foremost, we are looking for evidence of scholarly talent and achievement") and, on the same page, a naming-expecting department ("All applicants should be sure to list the faculty they hope to work with ... in the Statement of Purpose").
VERIFIED 2026-09-03 by direct fetch (desktop-UA curl, HTTP 200) with an exact-substring match on the quoted words after tag-stripping and whitespace normalization; no proxy, no cache. The social-science form of the spine ("well-conceived research interests, past research accomplishment, and intellectual biography") and the short-limit data point behind the rule that a complete statement written to a short departmental limit is never marked down for length: "Sociology program has a limit of 500 words (1 page) for the statement of purpose".
VERIFIED 2026-09-03 by direct fetch (desktop-UA curl, HTTP 200) with an exact-substring match on the quoted words after tag-stripping and whitespace normalization; no proxy, no cache. The signal-and-noise framing behind the space axis, in a social-science voice: "Send the best, clearest signal of your abilities as a future researcher, and minimize the noise around that signal."
Frequently Asked Questions
How long should my statement be?
Use the program’s stated limit. This rubric does not set one universal length, and a complete statement written for a short departmental limit should not be treated as inadequate for being short.
Do I need publications or a successful research result?
No. Research may come from supervised lab work, a course or thesis project, industry or field work, a self-run project, collaboration, or work done alongside a job. A null result, an unsuccessful attempt, or an unresolved problem can still be described as meaningful work.
Do I need to name a professor?
No. Naming a professor is one option, not a requirement. You can also explain how a group, facility, dataset, collection, course sequence, or research community would support the problem you want to work on.
Do I need a full research proposal or fixed plan?
No. You do not need a schedule, method plan, or fixed thesis proposal. Focus on clearly characterizing the problem you want to work on; it is fine to say your direction may develop or that you are open to related directions.
Get your essay scored across all 6 dimension with specific, actionable feedback. Two free reviews per day — no credit card required.
See how phd statement of purpose review works