'Which One Does Not Belong' and the Pull of Association
Mr. Rivera reviews eight picture classification sets with Ana (6 years 2 months). On the fruit set (apple, banana, carrot, orange) she picks the carrot: 'That's a vegetable.' On the animal set (cat, dog, cow, chair) she picks the chair. On the clothing set (boot, shoe, sock, glove) she picks the glove: 'Gloves are for your hands, not your feet.' On the tool set (hammer, saw, nail, screwdriver) she picks the saw: 'You could cut yourself.' Asked to look at the tool set again, she keeps the saw and adds, 'Nails are little.' She scores 4 of 8. Every set she gets right contains one picture from a wholly unrelated category; every set she misses contains an odd picture that is linked to the others by use, place, or a vivid personal association. Mr. Rivera must decide whether 4/8 can stand as a category-reasoning score.
1. Does Ana's 4/8 reflect weak category knowledge or a stable reasoning habit? Use her specific answers as evidence.
Read the worked answer
It is a reasoning habit, not a knowledge gap.
- She produces a correct superordinate label ('That's a vegetable') when the contrast is large, so she can form and name categories.
- Her misses share one structure: the odd picture is functionally or experientially tied to the rest. A glove belongs with footwear in her world because of where it is worn and stored; a saw belongs with tools because of what it does to you.
- On the tool set she does not re-categorize when prompted; she re-justifies, and 'Nails are little' is a physical feature, not a class rule. Repeating the same choice after a prompt is a strong sign that the association fires first and stops the search.
So the 4/8 measures something real but narrower than 'category reasoning': it measures classification when the distractor is distant. It should be reported with that qualifier.
2. Design one follow-up set that would separate 'cannot form the category' from 'forms a strong first association and stops.' Justify your choice in terms of the two competing answers.
Read the worked answer
Use a set where the associated item and the category-odd item are the same picture - no, better, where they conflict and the conflict is obvious once named:
hammer, saw, screwdriver, toolbox.
- Category answer: the toolbox, because the other three are tools and a toolbox is a container.
- Association answer: anything but the toolbox, because a toolbox is the most vivid, most 'tool-ish' item in the set.
Procedure: ask Ana to name the group first ('What are these three?'), then to choose. Two outcomes are diagnostic:
- If she names 'tools' and then still picks a tool, the category is available but association wins - a strategy problem, addressed by teaching 'name the group, then choose.'
- If she cannot name the group, or names it and cannot apply it, the category itself is the gap.
Complement it with four or five items of the same shape; one item alone is noise.
3. Ana's 4 of 8 is exactly the class median on this set. What may and may not be concluded from that?
Read the worked answer
May be concluded: she is performing like a typical child in her class on this particular mixture of items, which is weak evidence of relative standing.
May not be concluded:
- That her category reasoning is average. Her errors are patterned, not scattered; a child who missed different items for different reasons would carry a very different implication from the same total.
- That 4/8 is a stable estimate. On eight items each response is worth 12.5 percent; one lucky or unlucky choice moves the number noticeably.
- That the median is a meaningful benchmark when the item mix rewards large contrasts. If most of the class also defaults to association, the median is a description of the class, not a standard.
Report the pattern alongside the total: '4 of 8, with all errors occurring when the odd picture was linked to the others by use or place.'