Wellbeing & Assessment SuperSprint
Cambridge University Press & Assessment Internship | Research Sprint
Overview
This two-week SuperSprint brought together a cross-disciplinary team to explore how Cambridge English assessments affect test taker wellbeing and test anxiety. The goal was to identify opportunities to improve the assessment experience and support mental health.
The Challenge
Test anxiety affects up to 40% of candidates, potentially impacting millions of Cambridge English assessment candidates and disproportionately affecting vulnerable groups. With limited time, the sprint aimed to understand wellbeing across the assessment journey, identify where Cambridge could make a meaningful difference, and why this would be beneficial to all stakeholders and Cambridge itself.
My Role
I joined mid-sprint as a behavioural design specialist, I introduced a Cognitive Behavioural Model of Test Anxiety (von der Embse et al., 2013), chosen for its relevance to high-stakes assessments and testing with related users. While not specific to language tests, it offered a clear breakdown of anxiety predictors and their hierarchy of influence, providing a structured lens to interpret findings and compensate for the sprint’s limited reach.
I also supported interviews, led behavioural analysis, and helped develop actionable recommendations featured in the final sprint report.
The Process
1. Behavioural Framing
-
Secondary research: I conducted targeted secondary research to identify behavioural triggers of test anxiety, found a suitable model to help research.
-
Presented Model to team: I introduced the identified Cognitive Behavioural Model to the team to guide inquiry, focusing on two key predictors of test anxiety that became central in understanding user research:
-Self-efficacy: What lowers candidates’ perception of their confidence and preparedness?
-Social climate: What social influences affect the emotional environment around Cambridge tests and test preparation?
(Diagram of the Cognitive Behavioural Model of Test Anxiety (von der Embse et al., 2013) used, with my annotations and screen shot from the behaviour research feedback meeting)
2. User Research & Analysis
-
User Interviews: I supported interviews; taking notes, asking additional questions, and evaluated question effectiveness for emotional depth.
-
Set Themes: I worked collaboratively to generate themes for data tagging, using the model and research questions.
-
User Interview Analysis: I led interview transcript analysis, extracting key quotes, organising them on the team’s Miro board and tagging findings.
(Screen shots from interviews with key quotes and Miro board showing interview transcripts with tagged notes)
3. Analysis Workshop: Synthesis & Insight Generation
To translate interview data into actionable insights, I facilitated a collaborative affinity diagramming workshop with the team. The session was designed to be fast-paced and inclusive to non-designers.
-
Preparing the Data: Before the workshop we put the finding into high level groups structured around our pre-defined themes, to simplify sorting and guide focus for inexperienced team mates.
-
Affinity Diagramming Task: I guided the team through a structured, time-boxed affinity diagramming session, collaboratively sorting research notes to form patterns. Rapid insight generation was prioritised over strict adherence to the process, to suit the sprint’s pace.
-
Generating Insights: After forming patterns, I facilitated insight brainstorming followed by group discussion to formalise and prioritise insights.
Key Insights
-
Teachers strongly influence candidates’ sense of preparedness.
-
Candidates often feel isolated without peer support, with language test not necessarily taken as a cohort.
-
Mock tests were seen to be critical to determining student’s self-efficacy.
-
Exam dates were used as deadlines, leading to cramming rather than booking tests when ready.
These insights responded directly to the two key predictors, self-efficacy and social climate, highlighting how social support and preparation structures shape emotional readiness.
(Screenshots of affinity diagram and insights created in work shop, with connection to the behavioural model and links to initial recommendations)
The Solution
5. Forming Recommendations
Using the insights, we developed high-level recommendations focused on improving wellbeing and reducing test anxiety. The model was particularly useful in prioritising opportunities related to the key predictors of test anxiety:
-
Equip teachers to support wellbeing during exam prep, utilising their role in determining perceived preparedness.
-
Build candidate community to reduce isolation, give peer support and build confidence.
-
Investigate the emotional impact of mock tests on self-efficacy.
-
Research and add wellbeing guidance to pre-test materials to reduce cramming and promote readiness.
(Images of the creation of the sprint report presentation, showing concluding insights, opportunities and recommendations)
6. Deliverables
Alongside the team I helped compile the final report to record research for further use, my contribution focusing on behavioural and user research insights.
-
User Journey Visualisation:: I mapped interview finding across the assessment journey, highlighting mood fluctuations and influences. This was developed into a presentation tool for the report.
-
Insights and Opportunities: The insights we formed from user research where packaged into the report, concluding with our recorded next steps for improving wellbeing and continued research.
-
Quantified Impact: I returned to secondary research to help estimate the impact of test anxiety on Cambridge test takers, helping frame the scale of the issue persuasively.
Beyond the sprint, I was invited to support further behavioural research as it was a topic the team did not have much experience with. Including evaluating a gamification mode to advise on its use, and delivering an internal presentation on applying behavioural frameworks to UX design, to aide knowledge growth and hopefully supporting future projects.
(Images of user journey mood map iteration, from initial findings to final presentation for sprint report)