NURSE EDUCATOR BLOG

Strategies & Resources
for Nursing Education

 
 
 

NURSE EDUCATOR BLOG

State of Nursing Education

 
 
 

NURSE EDUCATOR BLOG

ATI Advocacy

Advancing Nursing Education Through Policy

 
 
Featured
State of Nursing Education
Nurse Educator Blog

State of Nursing Education

The State of Nursing Education is the authoritative ATI series examining key issues affecting today’s academic nursing programs, deans and faculty.

Explore All Articles

When Predictive Data Signals Risk, What Should Nurse Educators Do Next?

by  Michelle Perron     Nov 17, 2026
Header image 2026 Predictive Data MOF blog

 

 

Predictive data can provide helpful information about students before they progress to the more complex content that can derail them. These early warnings of attrition risk are useful, but only if the program acts on them.

For faculty and for program leaders, the challenge is to implement a consistent process for turning information about risk into analysis, support and follow-up across courses and cohorts.

This focused approach to student retention requires determining what the signal means, identifying where the student may need support, selecting an appropriate response or intervention, then monitoring whether that action produces improvement.

Research on learning analytics illustrates why these steps are necessary. Evidence suggests that predicting academic risk may be insufficient unless the finding is followed by a timely, well-designed intervention that encourages the student to engage.1

Determining which students are at risk is the essential first step. The difference maker is what a program does with information about that risk.

 

 

Treat a Risk Signal as a Starting Point, Not a Conclusion

A predictive score can draw warranted attention, but it cannot explain the factors influencing the student’s performance.

A student flagged at risk may be struggling with test taking, applying knowledge, poor study habits or a multitude of other issues. It’s that multitude that should give educators pause, because engagement and performance data may reflect circumstances not visible in a dashboard. Predictive information should prompt investigation, not labeling of the student.

Amy Edmison, DNP, RN, CNE, a former academic program leader who is now a Nursing Education Consultant for ATI, recommends that faculty ask four practical questions:

  • Where is the student performing well?
  • Where does the evidence show an opportunity for improvement?
  • Which resource addresses the specific need?
  • How will the student and faculty monitor progress?

This approach avoids turning a risk category into a conclusion about a student’s ability or potential by directing faculty back to the evidence and the decisions they can make based on that evidence.

Recent research reinforces the need for a broader view of student risk. A 2025 scoping review by Sheikoleslami and colleagues found that nursing student attrition is associated with intersecting academic, personal, social and institutional factors.2 The presence of these varied influences suggests that multifaceted approaches are more appropriate than single-response retention strategies.

Programs and faculty may need to look beyond the initial signal to consider:

  • Is the concern limited to one content area or present across several areas?
  • Is the student’s recent work an isolated setback or part of an emerging pattern?
  • When assessment results, course performance and learning resource use are combined, what do they show?
  • Is the student facing barriers the data cannot explain?
  • What additional information would help determine the best response?

Predictive data contributes to student success by drawing attention to risk. But it’s educator judgment that puts it into context.

Use Predictive Data to Start Earlier Conversations

Historically, faculty have identified struggling students through observation and by monitoring grades and assessments. This information unfolds over time, and some students get too far behind to recover.

Earlier information opens space and time for action. However, immediately assigning remediation isn’t always the best starting point. The best first response is usually a focused conversation.

Creating an environment that normalizes assistance and support makes these conversations less formidable. Reflecting on her program leadership experience, Dr. Edmison noted that students often became more comfortable asking for support when faculty and programs normalized struggle. “They didn’t feel like failures because they had to ask for extra help,” she said.

Laying that groundwork matters because students may not initiate conversations about academic concerns on their own. Predictive information can give faculty a reason to reach out before students withdraw as they see things spiral, or before grades drop too low.

The outreach should be specific and relational — not alarming. Consider starting the conversation by sharing what the data shows, Dr. Edmison said. Then ask the student what they think is contributing to the pattern. Discuss those issues together, then agree on a manageable next step.

A learning intervention study by Dai et al supports this connection between data and faculty communication.1 The researchers used predictive models to identify students at risk early in a course and then sent feedback designed to strengthen the student-instructor relationship. This feedback, provided in emails, included encouraging words along with the remediation instruction.

The study found that more than 30% of at-risk students accessed learning activities they had not previously used within 2 weeks of the feedback intervention. The researchers concluded that prediction by itself was insufficient, and that the practical value of student data depended on pairing identification with a timely feedback intervention.

Match Remediation to the Student's Identified Learning Need

After investigating the risk signal, faculty can turn their attention to the type of support the student needs.

To achieve remediation that is meaningful, Dr. Edmison recommends connecting assigned work to an identified need. The essential next step is providing opportunities for the student to apply the relevant knowledge and to receive feedback that clarifies what they should do differently.

Repeatedly assigning work that doesn’t address the student’s underlying difficulty is unlikely to increase their success. In the case of medical mathematics, for example, a student may need to connect the calculation with hands-on application. Instead of repeating the exercise, Dr. Edmison suggests asking them to explain the reasoning for the calculation or to identify when it would be required in a patient care situation.

This approach makes remediation more purposeful, with greater potential to address the student’s weaknesses.

 

 

Sheikoleslami’s research on nursing student retention points toward combined support strategies over isolated interventions.2 Academic assistance, mentoring, belonging, financial support and institutional responsiveness may all contribute, depending on the factors affecting a student’s progression.

Therefore, predictive data should not lead every student to the same remediation pathway. Instead, it should help faculty decide which evidence deserves closer attention and what type of response is most likely to be relevant and effective.

Monitor Performance and Adjust the Support Plan

As every nurse educator knows, completing a remediation plan doesn’t necessarily mean performance will improve. A student may finish an assigned tutorial, module or practice activity without resolving the weakness that prompted it. This should lead to reflections such as:

  • Did performance improve in the targeted weakness?
  • Can the student apply the remediation focus to a different question or in a different clinical context?
  • Did the student engage with the recommended resources?
  • Has the original pattern persisted, narrowed or expanded?
  • Does the student need additional feedback or a different learning approach?

Educators should consider changing course when a remediation plan is not working, Dr. Edmison said. “If they fully commit to a remediation plan and it doesn’t work, try something else.”

Here is where longitudinal data become especially useful. A single score captures a moment. Repeated measurement can help faculty determine whether a student is progressing, stuck, or developing difficulty in another area.

From a program standpoint, the value of monitoring extends beyond the individual student. At the cohort level, data provides broader perspectives that can identify trends. If multiple students repeatedly struggle with the same concept or NCLEX client needs area, faculty can examine the pattern and consider whether a curriculum issue or other factor should be addressed.

“Track and reflect on cohort data so that on evaluation days, you can brainstorm how to strengthen those areas with the students,” Dr. Edmison said. “Use your data to improve your program, which should improve your student experience as well.”

Give Students a Role in Understanding Their Progress

Indeed, the student experience determines the student outcome. That’s why students should be more than recipients of remediation. They can also be active participants in interpreting their progress and deciding what to do next.

Transparency is an important contributor. When appropriate, educators can help students understand what their performance data shows, where it is limited, and how it relates to the recommendations provided. Information sharing supports self-awareness and provides an opportunity for faculty to explain that the data reflects a probability, not a fixed outcome.

Supportive, forward-looking discussions can help a student understand:

  • What am I already doing well?
  • Where does my performance show a recurring difficulty?
  • Which learning activity or support resource should I use next?
  • How will I know whether the strategy is helping?
  • When should I meet with faculty again?

This type of sharing keeps the focus of remediation on growth and action rather than on the risk alone. It also supports responsible use of learning analytics, which is an expanding focus of attention.3,4 Research is showing that educator review and clear communication are important elements when analytics inform student support.

How ATI Pulse® Connects Predictive Data With Targeted Support

This process becomes more manageable when predictive information is connected with targeted learning recommendations and ongoing performance data.

To retain nursing students, identifying academic risk is an important first step. The next steps can be harder to take. Integrating systems that guide programs in responding to the data can assist.

ATI Pulse helps nursing programs move from performance information to student support decisions and action. Pulse uses assessment data from the ATI Content Mastery Series® (CMS) to predict a student’s likelihood of achieving proficiency on future proctored CMS assessments and their readiness for the NCLEX. These predictions can begin early in an academic program to help faculty initiate targeted remediation.

The predictive model in Pulse generates results based on data from more than 268,000 students across PN, ADN and BSN programs. It tracks student performance starting in the first semester and updates predictive scores as data becomes available.

Pulse provides personalized learning recommendations for each student’s identified areas of weakness. Throughout their academic journey, faculty can view student engagement with ATI learning resources and student performance in relation to national benchmarks.

This provides faculty with more than a warning signal. It also helps them:

  • identify which students may require closer attention
  • examine the areas contributing to the identified risk
  • connect students with aligned learning resources
  • review ongoing information as the student progresses
  • examine cohort-level patterns that may warrant broader discussion.

Pulse does not replace an educator’s expert assessment of student needs. Instead, it provides another layer of information that helps them make decisions and recommendations to keep students on track.

Early Insight Is Valuable When It Leads to Meaningful Action

Predictive data cannot provide a complete picture of a student, but it can provide information that helps programs understand their students better.

Today’s learning analytics can make an emerging concern visible while there is still an opportunity to act on it. Faculty can then provide the important human elements no algorithm can: context, judgment, conversation and encouragement.

Together, this delivers a student success approach rooted in data, education principles and human connection. Predictive data provides insights that create opportunities for action. Educators and students can then work together to turn these insights into purposeful next steps.

 

 

References

  1. Dai W, Lin J, Ji-Yoon Jin F, Tsai YS, Srivastava N, Le Bodic P, Gasevic D, Chen G. Learning analytics for Early Identification of At-Risk Students and Feedback Intervention. Journal of Learning Analytics. 2025;12(3):102-125. doi: https://doi.org/10.18608/jla.2025.8735
  2. Sheikoleslami RL, Princeton DM, Hansen LI, Kisa S, Goyal AR. Examining Factors Associated with Attrition, Strategies for Retention Among Undergraduate Nursing Students, and Identified Research Gaps: A Scoping Review. Nursing Reports. 2025;15:182. https://doi.org/10.3390/nursrep15060182
  3. O’Donoghue K. Learning Analytics within Higher Education: Autonomy, Beneficence and Non-maleficence. Journal of Academic Ethics. 2023;21:125-137. https://doi.org/10.1007/s10805-021-09444-y
  4. Tirado AM, Mulholland P, Fernandez M. Towards an Operational Responsible AI Framework for Learning Analytics in Higher Education. Knowledge Media Institute. 2024. https://doi.org/10.48550/arXiv.2410.05827

 

About the author: Michelle Perron is a healthcare writer and editor who develops evidence-based content for ATI Nursing Education. With more than 20 years of experience in healthcare publishing, journal development, medical editing, and editorial management, she helps connect nursing research and education best practices to today's faculty needs.