Prints on two pages in wide orientation. In the print dialog, choose "Save as PDF" to keep a copy.
Data-Driven Advocacy Guide: Reference Card
Class size, DFW contextualization, and data feminism1Planning Your Advocacy
Before gathering or presenting data, ask:
- What is happening? What is the exigence for this advocacy, and why now?
- What is your goal ? Which genre does your audience expect: a report, a dashboard, a one-page brief, a presentation?
- Who is your audience , and which data is legible and credible to them?
- Where did this data come from? Which biases and limitations should you acknowledge?
- Who is available to help with data collection and analysis : institutional research, the registrar, faculty governance?
- Which scholarship and national standards anchor your case?
- Are you making a comparison ? Which benchmark is appropriate: national, sector, or peer institution?
- What are the limitations of your data, and how will you name them?
- Which terms are legible to your audience: enrollment, student credit hours, retention rate, DFW rate?
- What is your circulation plan ? Who sees this, when, and through which channel?
2Class Size Case Study
Visualizing class size data for different audiences:
- For campus leaders: a dashboard showing average class size by department, overlaid with cost and student credit hour data. Present caps as a factor in retention as well as a labor issue.
- For faculty governance: class size distributions by course type and modality, with student learning outcomes beside cap size. Show the learning rationale.
- For public advocacy: a comparison of class size before and after a change, set beside retention rates or student satisfaction scores. Use the national distribution below.
- Equity check: do low-enrollment sections such as independent studies, honors sections, and internships distort your averages? Disaggregate by section type before computing means.
- Transparency rule: every visualization states its data source, its collection date, and its known limitations.
- Framing principle: present class size as a condition of student success, and cite the CCCC standard as disciplinary authority in place of personal preference.
Data Feminism Questions
Drawing on D'Ignazio and Klein, Data Feminism (2020):
- Who collected this data? For which purposes, and at whose expense?
- Whose goals are foregrounded: retention metrics, cost reduction, student learning?
- Who benefits from the current framing of this data, and who does not?
- What is not counted? Which working conditions, student needs, and structural factors are invisible in the metric?
- Who is harmed by the dominant interpretation of this data?
- Whose labor produced this data, and are they credited?
National Benchmarks
| Measure | Value |
|---|---|
| First-year composition, national median cap | 22 |
| Two-year college median cap | 25 |
| Four-year institution median cap | 22 |
| Institutions at 20 or lower | 36.5% (152 of 417) |
| Most common single cap | 25 |
| CCCC standard | 20 maximum, 15 preferred, 60 students per instructor per term |
Benchmarks: Community-sourced national course cap dataset, 435+ institutions, compiled by Holly Hassel and colleagues. Standard: CCCC, Principles for the Postsecondary Teaching of Writing (revised 2023). thewritingprogramexchange.org/data.html · Cole, K., Hassel, H., et al. (2025, revised 2026). Data-Driven Advocacy Guide. The Writing Program Exchange.
DFW Contextualization Guide
When an administrator raises DFW rates in composition courses| Step | What to do | Questions and breakdowns |
|---|---|---|
| 1. Slow Down and Contextualize Nationally | Ask first for the range of DFW rates in composition at institutions of your type. A 22 percent rate at an open-access community college differs from a 22 percent rate at a selective four-year institution. Your first move is a request for time to gather context. It is not a defense, a promise of fixes, or acceptance of the premise of a "high" rate. | |
| 2. Assess Your Student Support Infrastructure | Before attributing DFW rates to curriculum or instruction, ask what the institution provides. | → Which general academic support is available, and how do students reach it? → Which reading and literacy support exists? → Is the writing center adequately staffed, and is its outreach proactive or reactive? → Is library and research instruction integrated into courses? → Which support exists for multilingual students? → Does the institution connect students to support, or wait for self-referral? |
| 3. Disaggregate Your DFW Data | A single DFW rate hides variation. Request disaggregated data. See Derek Mueller's "Silhouette of DFWI" for approaches to visualization. | → By course type: 101, 102, developmental, co-requisite → By modality: face-to-face, online, hybrid → By student population: first-generation, Pell-eligible, multilingual → By section cap: 20 or lower, 21 to 25, 26 or higher → By instructor appointment: tenure-line, full-time lecturer, adjunct → By schedule and format: early sections, accelerated sessions |
| 4. Gather Faculty and Student Perspectives | Numbers alone do not explain what is happening. Gather qualitative data. | → Faculty survey: at the end of the term, ask instructors to explain anonymously why each student with a D, F, or W received the grade, and which resources were available for referral. → Faculty focus groups: which patterns do instructors see? Which support are they able to provide under current workloads? → Student survey: ask students who received DFW grades about barriers, about what would have helped, and about what the institution did not provide. |
| 5. Connect DFW to Faculty Working Conditions | Administrative conversations about DFW rates seldom include working conditions. Put the following questions to your own data before the meeting. | → How many writing students does each instructor teach per term, and how does the number compare with the CCCC maximum of 60? → How many instructors teach at more than one campus or institution, and how much time does each appointment leave for office hours and early intervention? → Does the DFW rate in your own data differ by section size or by instructor appointment type? |
Data feminism reminder
The DFW "problem" is often framed in ways which locate failure in students or instructors, and not in institutional systems, resource allocation, or working conditions. Ask who benefits from the framing and who does not. Naming the frame is the first step toward changing it (D'Ignazio and Klein, 2020).
thewritingprogramexchange.org/data.html · Cole, K., Hassel, H., et al. (2025, revised 2026). Data-Driven Advocacy Guide. The Writing Program Exchange.