Review, Methodology, and Representativeness of the CERA 2026 Spring Program Director Survey

The Council of Academic Family Medicine (CAFM) Educational Research Alliance (CERA) is an initiative to advance educational research in academic family medicine. CERA conducts recurring national surveys of academic family medicine faculty.¹ Members of the organizations associated with CAFM may submit up to 10 questions to be included in the survey. CERA’s general methodology has been previously described.² This report describes the survey development, administration, sample representativeness, and respondent demographics for the spring 2026 Program Directors Survey. Additional information is included, introducing a new rubric used in the peer review process. A new rubric was deemed necessary as the previous rubric failed to discriminate high-quality submissions for acceptance. The prior rubric provided an overall submission score between 1 and 5 with one opportunity for open-ended responses from the reviewer. This system led to a limited range and variability of scores. This gap made it difficult for the CERA survey directors to decide on accepted submissions.

A month-long call for submissions was opened in December 2025. Eighteen proposals were received for consideration. Each underwent independent review by three peer reviewers using a newly implemented scoring rubric. Ten reviewers (three CERA steering committee members, three former program directors, and four medical education researchers) participated. The rubric generated total scores ranging from 5 to 30 across five domains: background (1-5), hypothesis (1-5), significance (1-5), analysis plan (1-5), and question quality (2-10). The rubric also had three open-ended responses to elaborate on particular sections (background/hypothesis/significance, analysis plan, and question quality). Proposal scores and reviewer comments informed the survey director’s subsequent selection decisions.

Authors of accepted question sets (modules) were paired with CERA mentors to refine survey questions. After revisions, modules were submitted for inclusion, and a draft omnibus survey was pilot tested with family medicine educators outside the target population. Study teams incorporated pilot feedback before the CERA administrator programmed the final survey in SurveyMonkey (Symphony Technology Group, Menlo Park, CA). The final instrument included standardized demographic items³ followed by the accepted modules. The American Academy of Family Physicians Institutional Review Board approved the study in April 2026 (Protocol 19–366 ORG A35).

The sampling frame comprised all program directors of Accreditation Council for Graduate Medical Education (ACGME)-accredited US family medicine residency programs identified from a publicly-available ACGME list accessed in April 2026. No Canadian family medicine residency programs were included. The Association of Family Medicine Residency Directors (AFMRD) provided summary demographic data for potential respondents as of June 2026.

CERA distributed email invitations via SurveyMonkey. Four weekly reminder emails were sent to nonrespondents, followed by a final reminder 2 days before survey closure. Data collection occurred from April 28, 2026 to June 5, 2026.

We calculated descriptive statistics, including frequencies and percentages. χ² goodness-of-fit tests and Fisher exact test, as appropriate, were used to compare characteristics of potential versus actual respondents, excluding “no response” and “choose not to disclose” categories. Statistical significance was defined as α = 0.05 for all two-sided tests. Analyses were conducted using SPSS Version 31.0.0.0 (IBM, Armonk, NY).

The peer-review process resulted in five question modules selected for inclusion in the omnibus survey (Table 1).

The initial sampling frame included 724 residency program directors identified through the ACGME. Ten directors reported not yet graduating three resident classes, an eligibility requirement, yielding a final sampling frame of 714. A total of 262 directors completed the survey, for a response rate of 36.7% (262/714).

Demographic characteristics of respondents were compared with AFMRD membership data (Table 2). No substantive differences were observed by gender, race/ethnicity, or underrepresented in medicine (URiM) status, based on the Association of American Medical Colleges definition⁴; the Black, Indigenous, and People of Color definition; or self-reported URiM identification.

The spring 2026 CERA Program Directors Survey provides responses pertinent to family medicine education and is generally representative of the population based on AFMRD membership. This sample frame is the best available at this time, and it is still unknown if AFMRD membership is representative of all ACGME-accredited program directors. The response rate of 36.7% is the lowest CERA has had for this survey type. The range of prior response rates has fluctuated between 38.1% and 57.1%. The low response rate is a concern and follows trends in survey-based research across academic medicine.⁵ CERA has ongoing studies and interventions to attempt to reverse this trend. Efforts such as personalized invitations, population-specific marketing, and text-based invitations have been considered.

The implementation of the revised scoring rubric represents a meaningful change in CERA’s peer-review process. The expanded domains allowed CERA survey directors to differentiate proposals more effectively than the prior single-score system. This change may have downstream effects on the scholarly rigor of future CERA surveys by incentivizing clearer conceptual framing and more robust analytic planning among proposal authors. As additional cycles accumulate more observations, item-level analysis will help determine whether certain domains disproportionately influence acceptance decisions and whether weighting adjustments are warranted. From an administrative perspective, the new scoring rubric allowed for better discrimination between proposals to identify the sets of questions ultimately included in the survey.

CERA surveys continue to provide representative samples to survey program directors. CERA is taking active steps to continuously improve our internal processes, provide transparency of our methodology, and promote rigorous family medicine education research.

The authors acknowledge Ray Biggs, Rhiannon Johnson, and April Davies for their assistance in coordinating the technical aspects of survey creation and dissemination.

The authors declare no known or perceived conflicts of interest.