Clinical PhD Letters: Research or Practice?

Clinical PhD Letters: Research or Practice?

A clinical supervisor can be a useful recommender for a clinical psychology PhD, but should not automatically replace the person best placed to assess your research preparation. Start with the program's requested writer mix. Then identify what each proposed supervisor actually observed, including whether their clinical role involved research.

Our recommendation is to protect substantive research evidence in the reference set. Being excellent in a clinical workplace and being prepared for research training are related possibilities, not interchangeable conclusions. A letter should establish the connection through work the writer knows.

This guide concerns US clinical psychology PhD applications. PsyD, UK DClinPsy and medical-school references have their own requirements. The general PhD recommendation guide covers broader research-reference principles.

What two current programs actually ask for

Berkeley Psychology requires three letters and says at least two should be from professors or research supervisors. It welcomes close observations from academic, clinical, research or professional settings. Preserve that distinction between the required count and the recommended writer mix.

Northwestern Feinberg's Clinical Psychology PhD requests three letters from professors or supervisors familiar with academic, research and/or clinical work, preferably psychologists. It values deep knowledge of the applicant over distant familiarity. This is Feinberg's program, not a rule for all Northwestern psychology degrees.

These examples show why a universal answer such as “always use three professors” is inadequate. They also do not establish that three practice-only references are equally informative for every research doctorate. A permissible writer category is the starting point for choosing evidence.

For each target, record the count, required categories, preferences and any program-specific recommendation questions. Ask for clarification when a proposed arrangement falls outside published guidance. Another school's flexibility does not resolve the question.

Classify the observations, not the job titles

“Clinical supervisor” can describe very different relationships. One supervisor may know an applicant's professionalism during service work. Another may oversee recruitment, research assessments, protocol problems and analysis discussions. The title alone does not tell you which person can discuss research preparation.

Use this distinction when comparing candidate writers. Do not turn every competent clinical action into proof of research independence. Equally, do not discard a strong research witness simply because their title includes clinical work.

The employer recommendation guide covers professional references more generally. Here, the additional task is identifying which observations support the specific training you seek.

Constructed case: a strong clinical letter as the third perspective

The cases below are original illustrations, not admissions results.

An applicant has a thesis adviser, a research-project supervisor and a clinical workplace supervisor. The adviser observed the applicant formulate a question and revise an argument. The project supervisor saw data-quality decisions and collaborative research work. The clinical supervisor knows their reliability, boundaries and communication in a defined support role.

For a program with Berkeley's stated preference, this set provides two academic/research observers and a separate clinical perspective. The clinical letter's useful job is to describe the applicant's observed conduct. It does not need to imitate a research letter or claim that service experience proves the applicant can design independent studies.

Now suppose the applicant considers replacing the thesis adviser with a senior clinician who met them occasionally. The clinician has an impressive title but cannot describe more work than the existing clinical supervisor. Our recommendation would be to keep the substantive research witness, subject to the target's instructions.

The applicant can provide each writer with accurate context about the degree and the work they shared. They should not assign a desired percentile, dictate conclusions or ask all three writers to repeat the same account. Independent letters can complement one another without coordinated praise.

Constructed case: the clinical supervisor is also the research witness

A second applicant works as a coordinator in a clinical research setting. A psychologist supervises both their day-to-day responsibilities and their contribution to a study. A faculty lecturer knows the applicant mainly through a high course grade.

The supervisor observed the applicant identify an ambiguity in a coding procedure, document the problem, compare it with the written protocol and raise a question before making changes. Later, the applicant helped examine the consequences for an analysis and explain the limitation in a team report.

That is potentially useful research evidence. Its strength comes from the observed reasoning and the applicant's bounded contribution, not from calling the workplace a clinic. The supervisor should describe the actual responsibility and avoid suggesting that the applicant independently changed a protocol or made clinical decisions beyond their role.

Where the program accepts this writer and the complete set follows its guidance, our recommendation is to prefer that substantive observer over a lecturer who can only repeat a grade. Still check what the other letters contribute. One combined clinical/research relationship does not automatically cover every academic dimension the application needs.

The same distinction matters when the observer works in neuroscience or another related field. Our outside-field STEM reference guide explains how to connect relevant work without pretending the writer assessed an area they do not know.

Build the full reference set before choosing the final person

Applicants often frame the choice as a contest between two names. A better procedure is to compare complete sets. The best third writer depends on what the first two can already establish.

1. List the program's instructions. Separate mandatory categories from preferences and numerical limits.
2. Write one factual sentence per relationship. Identify the period, setting and work the person directly encountered.
3. Locate the research evidence. Which writer saw question formation, methodological reasoning, interpretation or sustained revision?
4. Look for duplication and absence. Three accounts of attendance leave a different gap from three accounts of one project.
5. Ask whether the writer can provide a substantive assessment. Give them room to describe the limits of their knowledge.

This is a planning exercise for the applicant. It is not a request to inspect confidential letters. You can know which projects a supervisor observed without knowing how they will evaluate you.

If a postdoc provided the closest daily research supervision while a PI supplied faculty oversight, consider the separate joint PI/postdoc letter question. Confirm that arrangement with the receiving program rather than assuming two signatures satisfy a particular category.

What if the supervisor cannot rate research potential?

Treat that as information about the relationship, not a problem to solve with stronger adjectives. The person may still provide valuable clinical observations if the program accepts them and the other letters supply the needed research perspective.

If the form asks for a comparison the writer cannot make, they should work from its actual wording and their real basis. The reference-rating guide distinguishes missing observation, a small comparison group and an actual reservation. Those are different situations.

A letter can explain the limits of a supervisory role while giving a clear account of what the writer did see. It should not infer research ability from clinical confidence alone or convert team achievements into individual accomplishments.

Give useful context without exposing people or writing the letter

Provide a current CV, the target instructions, relevant deadlines and a concise reminder of shared work. Describe a research task or professional responsibility at a level suitable for sharing. Patient identities and recognizable clinical stories are unnecessary for explaining your contribution.

The recommender chooses the assessment and its wording. If the person needs a factual correction about your dates or role, correct the supporting information you supplied. Do not write the evaluation in their voice or seek access to a confidential letter to make it match your statement.

Your own academic statement remains responsible for explaining your research interests and reasons for the program. A recommendation can corroborate preparation; it cannot answer a missing applicant prompt for you.

Review the narrative the writer controls

The academic recommendation rubric, available through recommendation letter review, addresses a supported faculty- or supervisor-authored narrative. It does not select the reference set, score a portal grid or assess clinical competence. Applicants can share the public guide with their recommender.

Our recommendation-letter collection covers related decisions. Choose the writer whose actual observations fill a real need in the complete application, while keeping the target's rules in view.

Primary guidance checked September 17, 2026. Cases and the observation table are original GradPilot illustrations.