Decision-making in programmatic assessment is only a challenge when we make it one.

Area of Science:

• Educational Assessment
• Learning Sciences
• Artificial Intelligence in Education

Background:

• High-stakes decisions in programmatic assessment for learning (PAL) are often perceived as intractable.
• Existing approaches may oversimplify competence as a single metric, neglecting the complexity of learning trajectories.

Purpose of the Study:

• To challenge the notion that programmatic assessment decisions are inherently intractable.
• To propose that decision difficulties signal modifiable implementation conditions.
• To advocate for a narrative approach to competence assessment, particularly in light of generative AI.

Main Methods:

• Critique of measurement paradigms focusing on single scores.
• Advocacy for a constructivist, narrative-based assessment approach.
• Identification of key institutional domains for realizing PAL's potential.

Main Results:

• The difficulty in programmatic assessment decisions often stems from underdeveloped narrative synthesis and lack of documented trajectories.
• Addressing these areas can significantly reduce decision-making complexity.
• Generative AI highlights the fragility of single-performance assessments, reinforcing the value of PAL.

Conclusions:

• Programmatic assessment, emphasizing narrative and documented progress, offers a robust response to the challenges posed by AI.
• Cultural transformation within five institutional domains is crucial for effective PAL implementation.
• Further research is needed on decision-making phenomenology and human-AI collaboration in assessment.