A useful way of assessing the coherence and consistence of your students’ work is the SOLO Taxonomy.

The Structure of the Observed Learning Outcome (SOLO) Taxonomy (Biggs & Collis, 1982) assesses student work, rather than skills or understandings. This is valuable as it emphasises the assessment of learning and outcomes over the assessment of personal characteristics that students might incorporate as a part of their identity (“I did badly at this maths task. That means I have more to learn” (growth mindset for learning) versus “I don’t have the skills or understandings to do maths. I’m not a ‘maths person’.” (fixed mindset for learning)).

The SOLO Taxonomy qualitatively describes student work as sitting within one of five phases of increasing complexity, or transitioning between two of these.

The phases include:

  • Prestructural, in which there is no relevance or coherence between elements;
  • Unistructural, in which a single relevant idea is focused upon;
  • Multistructural, in which multiple ideas are generalised in a conclusion, additively;
  • Relational, in which multiple related ideas lead to a credible and supported conclusion; and
  • Extended Abstract, in which multiple related ideas are extended to resolve inconsistencies, and conclusions are held open or qualified to allow logically possible alternatives.

Depending on the topic and learning area, and the amount of rehearsal students have had of the skills required by the task, students at various stages of their learning might provide assessment pieces that could be assigned any of these stages. The idea behind a unit of work is to move students from prestructural phase to the relational or extended abstract phase.

Visual representations of the stages of the SOLO Taxonomy
Visual representations of the stages of the SOLO Taxonomy

Example: Why does the moon appear to change shape?

Imagine we are working with students in scientific inquiry about why the moon appears to change shape. Student responses to this inquiry question can be assigned to the five phases. Examples are below.

Prestructural. A prestructural response to this question might be anything from “I dunno” to “because it changes shape” (tautological response) or “it looks different every day.” Such responses don’t reveal any understanding of the phenomenon. Prestructural responses might also include denial, for example “it doesn’t look like it changes shape to me“. Prestructural responses contain a mismatch between the cue and response, are inconsistent or contradictory, and reveal no understanding of the topic.

Unistructural. A unistructural response to this question focuses on one relevant aspect with limited generalisability. For example, the student might respond with “the moon doesn’t really change shape” or “the moon reflects light from the sun“. These responses both give part of the answer but not a complete answer to the question.

Multistructural. A multistructural response would detail several aspects or elements, but each are separate and together they are additive. For example, a student might say:

I’ve seen the moon change shape. Some days it’s whole and other days it’s small. The moon reflects light from the sun. It’s in a different position all the time.

This response contains several elements of an explanation for why the moon appears to change shape, but the sentences don’t really relate to each other and they don’t develop a complex explanation for this phenomenon.

Relational. For example, a student might respond to the question about the moon by suggesting:

The moon doesn’t really change shape, it just looks like it does. The sun shines on the moon, just as it does on the Earth. As the positions of the sun, moon and Earth shift, the amount of the moon that we on Earth can see lit up by the sun waxes and wanes. When we can see the whole face of the moon that is fully lit, we see a full moon. When we can’t see any part of the illuminated side of the moon, we call it a new moon.

A relational response includes several relevant ideas integrated into a coherent whole. Links are identified. There is consistency. A relational response might include a comparison, or inductive reasoning.

Extended abstract. An example extended abstract response to why the moon appears to change shape is:

The moon appears to change shape because of the changing positions of the sun, Earth and moon. The Earth orbits the sun, and the moon orbits the Earth. I observed the moon over three months and noticed that it takes about 29 days to orbit the Earth. The half of the moon that faces the sun always receives and reflects the sun’s light. When the moon is on the far side of the Earth from the sun, we can see most or all of this lit half of the moon; a waxing gibbous moon, a full moon, or a waning gibbous moon. When the moon is between the sun and the Earth, we see less than half or none of the lit half of the moon; a waning crescent moon, new moon, or waxing crescent moon. The moon continues in this cycle. But how come we don’t see a lunar and solar eclipse every month? This is because the sun, Earth and moon are not on a single plane. The moon’s orbit around the Earth is on a plane that is ~5 degrees tilted from the plane of the Earth’s orbit around the sun. This is just enough for solar and lunar eclipses to be avoided most of the time.

An extended abstract response consists of an integrated whole conceptualised at the highest level and generalised to a new topic or area. The author is able to generalise and hypothesise and identifies where they have done so. Inconsistencies are identified and resolved. Alternatives are considered, and conclusions held open pending qualification. Deduction and/or induction are used to defend the thesis.

You can see that the level of complexity of an extended abstract response is significantly more sophisticated than even a multistructural response. Such a response requires comprehensive and deep understanding of the topic, as well as rehearsal of the skills for communicating it, and a knowledge of the criteria for assessment of the response.

Have you used the SOLO Taxonomy to assess your students’ work?

Have you any questions about how the SOLO Taxonomy could be used in the primary or middle years classroom?


Biggs, J. B., & Collis, K. F. (1982). Evaluating the quality of learning. New York: Academic Press.