Solo Taxonomy

Theory: SOLO Taxonomy | Template: The Confession | Words: 1,867

# SOLO Taxonomy: Beyond Bloom's Cognitive Domain

For a long time, the field of education and learning design operated under a fairly straightforward assumption: Bloom's Taxonomy provided the ultimate ladder for understanding and assessing cognitive development. It was our North Star for everything from curriculum planning to grading. We believed that by moving students from 'remember' to 'understand' and eventually to 'create', we were capturing the full spectrum of intellectual growth. But a deeper look, fueled by decades of research and practical application, has fundamentally changed how we approach evaluating true understanding. We've realized that Bloom tells us what a student did, but it doesn't fully tell us how well they did it structurally.

What We Used to Believe

We used to believe that Bloom's Taxonomy, with its hierarchical levels ranging from remembering facts to creating new ideas, offered a complete framework for assessing learning. It felt intuitive. A student could recall information, then explain it, then apply it in a new situation, then break it down, evaluate it, and finally, build something original. This progression seemed to cover all bases. Educators across disciplines adopted it as the standard for writing learning objectives, designing classroom activities, and crafting rubrics. It provided a common language for describing intellectual tasks.

The power of Bloom's was its clarity in defining cognitive operations. It helped us articulate what we wanted students to do with information. If a task asked students to "analyse" a text, we understood that they needed to go beyond summarizing and instead identify components and relationships. If they were asked to "evaluate," they needed to make judgments based on criteria. This focus on the verb, on the action, was incredibly helpful for structuring instruction and assessment. It became the default lens through which we viewed learning quality, assuming that achieving a higher Bloom's level inherently meant a deeper, more sophisticated understanding.

The Turning Point

The first cracks in this seemingly complete framework started to appear not because Bloom’s was wrong, but because it was incomplete. We began to observe situations where students performed a task at a high Bloom's level, like "applying" a concept, but the quality of that application varied wildly. Imagine two students asked to apply a scientific principle to solve a new problem. Both might successfully "apply" the principle. However, one student might haphazardly throw a few relevant facts at the problem, while the other might systematically integrate multiple concepts, explain their interrelationships, and construct a robust, coherent solution.

The cognitive operation – "apply" – was the same for both students. Yet, their responses demonstrated vastly different levels of structural sophistication and depth of understanding. Bloom's didn't give us the language to differentiate this crucial distinction. It could tell us they both applied, but not how well they structured their application. We saw this pattern repeatedly: a student could list several correct points (high volume) but fail to connect them into a meaningful whole (low structure). This highlighted a critical gap in our assessment toolkit. We needed a way to measure not just what cognitive action was performed, but how the knowledge was organized and integrated within the response itself.

The Research That Changed Everything

This growing awareness of Bloom's incompleteness paved the way for a deeper understanding of learning quality, largely driven by the work of John Biggs and Kevin Collis. Their seminal work, Evaluating the Quality of Learning: The SOLO Taxonomy (Biggs & Collis, 1982), introduced a framework that fundamentally shifted our perspective. SOLO, which stands for Structure of the Observed Learning Outcome, offered a way to assess the complexity of a student's response, moving beyond a simple right or wrong answer to evaluate the depth of their understanding.

SOLO identifies five distinct levels of structural complexity:

1. Pre-structural: The student misses the point entirely or gives irrelevant information. They haven't engaged with the task meaningfully. 2. Uni-structural: The student focuses on one single, obvious aspect or idea. They understand only one piece of the puzzle. 3. Multi-structural: The student identifies several relevant aspects or ideas, but they are treated as separate, unconnected pieces of information. They have multiple puzzle pieces, but haven't put them together. 4. Relational: The student integrates several relevant aspects into a coherent structure. They understand how the different pieces of information relate to each other and form a meaningful whole. This is where true understanding of connections begins. 5. Extended Abstract: The student not only integrates the given information but also generalizes it to new contexts, makes predictions, or raises new questions beyond the original problem. They can take the completed puzzle and see how its principles apply to other puzzles.

This hierarchical structure of SOLO has been empirically validated across different subjects and educational levels, demonstrating its reliability in differentiating understanding (Chan, 2004; Yates & Masters, 2011). It provides a robust measure of how students build and connect knowledge.

The practical implications of SOLO quickly became evident. This is because SOLO allows educators to pinpoint exactly where a student's understanding breaks down structurally, guiding them on how to integrate their knowledge more effectively. As Hattie and Brown (2004) emphasized in their meta-analysis, effective feedback is crucial for learning, and SOLO provides a powerful framework for structuring that feedback to target specific levels of understanding.

Furthermore, SOLO aligns beautifully with the concept of constructive alignment, where learning activities and assessment tasks are explicitly designed to match intended learning outcomes (Biggs, 1999). By using SOLO, we can define the desired complexity of those outcomes, ensuring that our teaching and assessment truly reflect the depth of understanding we aim for. For instance, in engineering education, SOLO has been successfully used to assess the sophistication of student design solutions, showing its versatility beyond traditional academic subjects (Smith et al., 2005).

What the Evidence Shows Now

The evidence now clearly shows that Bloom's Taxonomy and SOLO Taxonomy are not interchangeable, nor is one superior to the other. Instead, they are complementary, orthogonal dimensions of learning assessment. Bloom classifies the cognitive operation a student performs – are they remembering, applying, or creating? SOLO, on the other hand, classifies the structural complexity of the student's response – how many elements did they integrate, and how well did they connect them?

Consider the example from our earlier discussion: a student is asked to "Apply" a concept (Bloom Level 3). One student might apply it in a multi-structural way, listing several relevant steps without clearly linking them to the outcome. Another student might apply it relationally, integrating those steps into a coherent, logical sequence that demonstrates a deeper understanding of the underlying principles. Same Bloom level, but profoundly different SOLO levels, reflecting different qualities of learning.

This distinction has profound implications for how we design and evaluate learning experiences. It moves us beyond simply asking "did they get it right?" to "how well did they construct their understanding?"

Real-world applications reinforce this. The University of Auckland, for example, redesigned assessment tasks in a postgraduate course using the SOLO framework. This focus on aligning learning outcomes, teaching activities, and assessment criteria led to an 8% increase in the average grade on the redesigned tasks, alongside improved student understanding of expectations (University of Auckland, 2018). Similarly, a pilot program in Queensland schools demonstrated a 15% improvement in writing scores for students who received SOLO-based feedback, which focused on the structural complexity of their arguments (Queensland Department of Education, 2022). These results highlight the tangible benefits of moving beyond a singular focus on cognitive operation to also consider structural quality. This shows a direct impact on fostering deeper learning.

This dual perspective is especially critical in the age of AI-driven learning. While AI can readily assess if a student has performed a certain Bloom's operation – identifying keywords for "remembering," or checking steps for "applying" – it often struggles with the nuanced evaluation of structural coherence and integration that SOLO provides. AI might count relevant points (multi-structural), but discerning true relational understanding or extended abstract thinking still requires a framework like SOLO, often guided by human expertise or sophisticated AI trained with SOLO-informed rubrics.

The Framework

Integrating Bloom's and SOLO offers a powerful, nuanced approach to learning design and assessment. Here’s a practical framework for how we can apply this dual lens:

1. Design Tasks with Bloom: Start by clearly defining the cognitive operation you want students to perform. Are they analyzing? Synthesizing? Evaluating? This sets the initial expectation for the type of thinking required. For example, "Analyze the economic impacts of climate change." 2. Define SOLO Levels for Each Bloom Task: Once the Bloom level is set, articulate what a uni-structural, multi-structural, relational, and extended abstract response would look like for that specific task. For our climate change example:

• Uni-structural: Mentions one economic impact, like rising insurance costs.
• Multi-structural: Lists several impacts (insurance, agriculture, tourism) but doesn't connect them.
• Relational: Explains how rising insurance costs impact tourism, which then affects local economies, integrating these elements.
• Extended Abstract: Generalizes these impacts to global supply chains or proposes new policy frameworks beyond the immediate scope.

3. Craft Rubrics with Both Dimensions: Your rubrics should explicitly address both the Bloom's cognitive operation and the expected SOLO structural complexity. This makes assessment criteria transparent and actionable for students. Practical guides like Guide to the SOLO Taxonomy by Hook and Mills (2011) offer excellent examples for rubric development. 4. Provide Targeted Feedback: Use the SOLO levels to give students specific feedback on the structure of their learning. Instead of just saying "your analysis is weak," you can say, "You've identified several impacts (multi-structural), but how do these impacts connect and influence each other to form a cohesive argument (relational)?" This type of formative assessment is key to guiding students towards higher levels of understanding (Stiggins et al., 2006).

This framework ensures that we're not just measuring if students can perform a cognitive trick, but if they can truly build, connect, and generalize knowledge in a meaningful way.

The Invitation

This journey from a singular, widely accepted framework to a more nuanced, dual-lens approach has been profoundly impactful for how we think about learning and assessment. It challenges us to look beyond surface-level performance and truly understand the architecture of a learner's knowledge. The shift from seeing Bloom as the be-all and end-all to recognizing the indispensable role of SOLO has opened up richer conversations about quality, depth, and the very nature of understanding.

How can we best leverage SOLO to assess AI-driven learning?