Reading focus
Syllabus: How Do We Describe and Measure Individual Differences?
Trace the move from positive correlations among cognitive tests to a hierarchy of abilities, keeping broad and specific differences visible alongside the general factor.
Why you’re reading it. See what a well-established tradition has built from test scores, while keeping its measurement model separate from claims about causes and outcomes.
A question to consider. If the learners’ vocabulary, spatial, and reasoning scores correlate positively, what would a general factor summarize, and what differences between learners would it leave out?
Why this reading
This primer introduces the study of differences in cognitive performance. It connects the tendency for cognitive tests to correlate positively with a hierarchy of general, broad, and specific abilities, then places measurement alongside questions about development, biological explanation, and life outcomes.
Before you read
A test battery combines several tasks. The positive manifold is the tendency for performance on different cognitive tests to correlate positively across people. A factor summarizes shared variation under a statistical model. General intelligence, or g, refers to the common component; it does not exhaust the differences among people or the skills a particular task requires.
Read for
- Distinguish the observed pattern of correlations from the hierarchical model used to represent it and from proposed explanations of that pattern.
- Follow how broad and specific abilities remain relevant alongside a general factor.
- Separate evidence about measurement, prediction, and causes when the primer moves between research questions.
Consider afterwards
- Two people obtain similar overall cognitive scores but have different patterns of verbal and spatial performance. What would a hierarchical description preserve that a single score conceals?
Editorial context
This is a compact account of a research tradition, not a complete survey of competing theories. A statistical general factor does not by itself identify one biological cause. The mutualism model, for example, proposes that interacting processes can generate positive correlations. The genetics discussion reflects 2013: its statement that genome-wide studies had not identified additional significant variants has been superseded by later large-scale work. Associations within a population do not determine the causes of differences between populations.