The student model attempts to create a computational representation of particular aspects of the student, for a specific purpose. Most often, the student model represents qualities such as knowledge level, personality traits and preferences, which are reasoned upon to guide the instruction by the [[Intelligent tutoring system|intelligent tutoring system]]. It often goes hand-in-hand with personalisation, in that a student model guides the system to make decisions in line with their understanding of the student. However, **a student model is not necessary for [[Personalisation|personalisation]]**. For example, we can present three different math worksheets with differing difficulties, where the student can make a self-judged choice about which is appropriate, thereby personalising their learning. The need to understand the student and create a computational representation of pedagogy that can reason upon this student model, can lead to learning designers constraining and reducing education; which is often a criticism of [[Intelligent tutoring system|intelligent tutoring systems]]. In contrast, theories like [[Constructionism|constructionism]] often let go of the moment by moment control and expect that through the activity of constructing a physical artefact that relevant learning primitives will be triggered, thereby not requiring a student model. Increasingly student modelling is being used as a blueprint to create [[Simulated student|simulated students]] that are used for teacher training, evaluation of systems, and research. --- The student model can consider a variety of traits, that I have just started organising loosely in the following where there are overlapping dimensions. As usual with making sense of the complexity of the human mind, there are always reductive decisions. In particular, there is a tradeoff between the extent of reductionism and the pragmatics of ease of derivation or use in a computationally friendly format. Whilst here I categorise and denote it linearly as limited by the modality of writing, the true representations of relationships will form complex interconnected webs. **Traditional knowledge** * Mastery over a set of [[Knowledge component|knowledge components]] * Misconceptions * Skill proficiency **Cognitive traits** * Learning rate * Forgetting rate * Reasoning ability **Metacognitive traits** * Self-efficacy * Help-seeking behaviour * Goal orientation **Emotional state** * Frustration (links with modelling [[Productive failure|productive failure]]) * Boredom * [[Flow]] * Motivation **Personality traits** * [[Big five personality trait model]] **Demographics** * Age (links with [[Piaget's theory of cognitive development|the stages of cognitive development]]) * Family background * Gender --- The **behavioural traits** of the student are the qualities that we can observe. The specific behaviours can give us insight into some of the higher-order traits described above. Also vice-versa, the higher-order traits of the student can aid in inferring the student behaviours. Though the possible student behaviours are constrained and shaped by the [[Mediation|mediating]] tool. For example, the behavioural interaction with a video includes pausing, rewinding, increasing speed and skipping. Whilst interactions with a [[Dialogue tutoring|dialogue tutor]] provides flexibility through natural language interactions; utterances still conform to underlying patterns [[@graesserCollaborativeDialoguePatterns1995|(Graesser et al. 1995)]]. --- # References Graesser, A. C., Person, N. K., & Magliano, J. P. (1995). Collaborative dialogue patterns in naturalistic one-to-one tutoring. _Applied Cognitive Psychology_, _9_(6), 495–522. [https://doi.org/10.1002/acp.2350090604](https://doi.org/10.1002/acp.2350090604)