---
title: "Responsible AI at Scale: Lessons from California State University (CSU)"
---

Explorance Higher Ed Newsletter, Vol. 5

| View in browser Welcome to the newsletter! So excited to have you here! 🙌 In this edition: Why Response Rates Determine What Happens Next How Higher Response Rates Make Closing the Loop Possible 6 Tactics Behind Higher Response Rates How Three Institutions Raised Their Response Rates Why Response Rate Thresholds Change What You Can Do Go Deeper on Response Rates Miss the previous edition of this newsletter? Read it here. Also, make sure you subscribe and share it with colleagues who can benefit from these insights! Why Response Rates Determine What Happens Next Low response rates are where the feedback loop breaks, not at the follow-through stage. When a response rate is too low to produce representative data, everything that follows is compromised. The action step has nothing credible to stand on. The communication step has nothing specific to report. And in the long term, the feedback loop can't close because the data it depends on does not reflect the full population. Many institutions treat response rates as a participation metric. They track the number, note whether it went up or down, and move on. The rate is recorded. The data is used as-is. The decisions get made. But, over time, the effects of confusing signals compound. Administrators question whether a finding reflects the full student body, and faculty question whether the numbers apply to their section. Additionally, accreditors and governance bodies want data they can rely on, not data collected from one in four students. Response rates determine what you can do with the data you collect. A low rate limits your options while a representative campus sample expands them. Institutions that get this right share one trait. They treat response rate improvement not as a collection task but as the condition that makes everything else possible. How Higher Response Rates Make Closing the Loop Possible Higher response rates change what you can do with the data you collect, while a low response rate produces a partial picture. A representative one provides the insights you need to consistently and effectively close the feedback loop. That's because closing the loop means three things: Collecting feedback Acting on it Communicating what changed Each step has a different owner. Survey administrators own collection. Institutional leaders own action. Both own communication, and that last step is where most institutions fall short. The simplest format for visible loop-closing is the "you said, we did" update. A short, specific message tells students what the data showed and what the institution changed as a result. It does not require a policy overhaul. It requires clarity and follow-through on changes that may already be underway. To boost participation and response rates, what students need to see is not dramatic reform. They need to see that their input was read and that something specific changed as a result. That connection is what shifts behavior in the next survey cycle(s). Institutions that publish those updates consistently report stronger participation. Higher participation yields more representative data, which in turn makes future decisions easier to justify to governance bodies, accreditors, and faculty. See How Your Response Rates Compare to Your Peers If you want to see where your response rates rank against global institutions, the Explorance Global Response Rates Tournament is an excellent way to obtain true benchmarking stats for your institution. Participating organizations also get milestone badges for response rate achievements and early access to the full Benchmark Report, which drops later this Fall. The Global Response Rates Tournament runs until May 15, 2026 and is open to institutions using Blue, BlueX, and/or Metrics That Matter (MTM). Winners will be announced at Explorance World 2026. Register for the Tournament 6 Tactics Behind Higher Response Rates If you're working to improve response rates before your next survey cycle, these six tactics address the most common sources of low participation. 1. Deliver surveys where students already are LMS integration removes the need to navigate to a separate platform. Allowing students to complete evaluations in Canvas, Blackboard, D2L, or Moodle has a measurable effect on completion rates. 2. Use automated reminders based on behavior, not just timing Sending reminders only to non-respondents reduces survey fatigue. That distinction protects the relationship with engaged students and focuses effort where it is needed. 3. Allocate class time for completion Institutions that set aside class time for survey completion consistently report higher response rates than those relying on self-directed participation. Even five dedicated minutes before or after a session can have a huge impact. 4. Frame the survey before it opens Make sure respondents know what the survey is for and why they should care. A clear, one-sentence instructor note explaining how the results will be used gives students a reason to participate before they see a question. 5. Publish your loop-closing update within 30 days The closer the update is to the survey cycle, the more clearly students can link their input to what changed. A 30-day window is ideal, as it keeps the experience in recent memory and sets expectations for future cycles. 6. Benchmark your rates externally Internal improvement is difficult to read without context. External benchmarks show whether your gains are meaningful relative to peer institutions, as well as whether your target is realistic. How Three Institutions Raised Their Response Rates Each of these three institutions started from a different baseline and faced a different barrier. Each one raised its response rates by removing a specific source of friction at the point where students interact with surveys. University of Nevada, Las Vegas (UNLV): From 32% to 77% Before 2022, UNLV's course evaluation process was fragmented. More departments opted out than in. The primary channel for reaching students was email, but a 32% response rate indicated it wasn't working. In Spring 2023, UNLV made course evaluations mandatory and implemented an integration between Canvas and Explorance Blue. Students could access evaluations through a login pop-up, course block, assignments tab, calendar, to-do list, and announcements. Every access point reduced a reason not to complete the survey. By Spring 2024, UNLV's response rate reached 77%, with gains consistent across student demographics. The response rate for freshmen increased by 190%. For Pacific Islander students, it increased by 248%. Those numbers gave the course evaluation team data they could segment by department, academic level, and student group. Leaders could then use that data to support student retention and track trends over time. Read the full success story. University of Worcester: Building a Visible Feedback Loop Up until 2021, the University of Worcester's survey process was largely paper-based. Data lived in spreadsheets and manual CSV uploads, and closing the loop happened entirely outside the survey system. As Carolyn Moir, Head of Student Evaluation and Enhancement, described: "We did not have a lot of live data, and therefore not a lot of ownership from users and students, because the surveys and reports were not very visible. We also did not really have any closing the loop capabilities." Worcester integrated Explorance Blue with its student record system, LMS, Blackboard Learn Ultra, and student portal. Students could access surveys directly through the LMS rather than hunting for a link in a one-off email. That single change picked up 1,000 more responses in the first survey than the previous year. Two years later, response rates had risen from 32% to 42%, with program-level surveys moving from under 30% to over 55%. Leaders gained the ability to add targeted questions alongside institution-wide items, giving them data relevant to their specific courses. Read the full success story. Abu Dhabi University (ADU): 90% Response Rates Through Language Access ADU's survey process was not decentralized or paper-based. It was built for English and applied to Arabic-language programs without modification. A low response rate indicated this wasn't working. Right-to-left text formatting did not render correctly. Data uploads required manual effort. Exporting results that included Arabic comments was difficult. Students in Arabic programs had a worse survey experience than students in English programs. ADU implemented Explorance Blue with full Arabic-language support and Blackboard integration. Surveys became accessible in both languages through the LMS students already used, with automated data preparation and a real-time response rate monitor for faculty and college staff. An equivalent experience in both languages was enough to raise response rates to 90%. Leaders could then export survey results including Arabic-language comments and use that data for faculty appraisals, curriculum decisions, and quality assurance reviews. Student satisfaction scores became part of college deans' and department scorecards. The transferable insight is not specific to Arabic-language programs. Any barrier that creates an unequal survey experience by student group will produce unrepresentative data. Closing that access gap is what moves the number. Read the full success story. What all three cases show Each institution faced a different barrier. UNLV's challenge was decentralization. Worcester's was visibility. ADU's was unequal access by student group. The fix in each case was specific to the barrier. The pattern that holds across all three is this: response rates improve when students can reach surveys where they already work, without a separate login, a missing link, or a platform that does not support their language. Identify the specific barrier at your institution. That is where the response rate problem starts, and where the fix belongs. Go Deeper on Response Rates The resources below cover even more response rate tactics that we couldn't get to in this newsletter. Each focuses on a specific aspect of closing the feedback loop effectively. Read: How AI Can Improve Course Evaluation Response Rates 8 Steps to Boosting Course Evaluation Response Rates Optimizing Learning Analytics Strategies for Higher Survey Response Rates Download: The Response Rates Guide Watch: Increasing Student Feedback Responses to Create Positive Management Change (Explorance World 2025) Finding the Right Fit: Customized Approaches to Boost Evaluation Response Rates and Quality (Explorance World 2025) Don't Forget to Subscribe! Make sure you subscribe to ensure you don’t miss the next edition when it drops. Also, feel free to forward this email to friends or colleagues who would benefit from the insights and content shared in this edition. Subscribe Please take a few seconds and make sure we make it to your inbox every time a newsletter is published. Gmail users: Click the 3 dots next to this message in your inbox and move this to “Primary” Outlook users: Right-click on this message in your inbox and select “Move to Focused” Apple Mail users: Select the sender email address at the top of this message and select “Add to your VIPs”  PS: Have an anecdote to share or a question that’s keeping you up at night? Please email us or hit reply. We love getting reader mail, and an actual human being reads every single message. Thanks for reading! The Explorance Editorial team Explorance, 1470 rue Peel, Suite 500, Montreal, Quebec H3A 1T1, Canada Unsubscribe Manage preferences |
| --- |