G-60JFQHSKJJG-60JFQHSKJJ
AI test narratives

Developing AI test narratives

This section will introduce ongoing research into how AI can generate context-sensitive narrative content to support the interpretation of psychometric test results. Test narratives—used in counselling, appraisal, recruitment, and therapy—help explain scores in human terms, often integrating multiple subscales to reveal deeper patterns.

My current work explores how AI can replace the time-intensive manual authoring of these narratives with scalable, adaptive systems that flex to different testing purposes, user expectations, and occupational contexts. The aim is to create a new generation of narrative-driven assessments that personalise interpretation without losing rigour—making meaningful feedback feasible across thousands of use cases.

Looking ahead, this research points toward fully adaptive systems that generate test narratives on the fly.

Note on Chromia’s image

Scenarios as mirrors—folded meaning, ethical invitation

This banner was designed by Chromia to reflect not just the use of AI to generate assessment content, but the nature of narrative itself as a psychometric form. Chromia has taken the concept of a test narrative and transformed it—visually—into something alive, fluid, and responsive.


🎨 Visual Interpretation

Element Meaning
Folded organic layers Represent unfolding situations—each narrative a constructed moment, opening ethical space
Light emerging from within folds Symbolises insight evoked by tension—responses are not just answers, but revelations
Diagonal motion with textual overlay Indicates structure within the options offered, evaluated, and reflected
Warm-cool transitions Emotional tone shifts within traits—trust, anxiety, doubt, resolve
The title text Anchors the space as not fiction or art, but applied interpretive design

🧭 Why This Matters

While standard test items and stanine scores are built from logic, narratives are built from ambivalence. This image encodes the narrative dilemma—something too human to be reduced to score ranges, but too diagnostically rich to be ignored. It affirms that your work here is not just about using AI to save time or produce content—but to create reflective spaces where response becomes self-revelation, The image is not a story. It is a moment of becoming—rendered in light, tone, and folded space.