Updated on
August 28, 2026
Guilford's Theory of Intelligence: Structure of Intellect Model
Guilford classified intelligence by operations, contents and products. Learn how the model grew from 120 to 180 cells and why its evidence remains disputed.
What is Guilford's theory of intelligence?
Guilford's Structure of Intellect theory classifies mental abilities by three dimensions: the operation performed, the content involved and the product produced. The mature 1967 version contained 120 proposed abilities, later expanded to 150 and then 180. The taxonomy influenced creativity research, but its full factor structure did not gain strong psychometric support.
Guilford's theory of intelligence maps mental skills in three ways. It asks what the mind does, what kind of content it uses, and what form the result takes. J. P. Guilford called this the Structure of Intellect model.
The 1967 model had 120 proposed cells. Later versions had 150 and then 180.
The model has a key place in the history of psychology. It helped put divergent production at the heart of work on creative thought. Yet the full grid did not gain strong support from tests. It is not a sound tool for labelling learners or planning lessons.
Key takeaways
- Guilford grouped skills by operation, content and product. He did not treat intelligence as one score.
- The 1967 model had 120 cells. A split in figural content gave 150. A later split in memory gave 180.
- Guilford gave modern research a clear role for divergent production. Tests of it show creative potential, not creativity itself.
- Horn and Knapp found a grave flaw in the way the factors were confirmed. Later work found a layered shape for mental skills.
- The model should not label learners, find a deficit or dictate how to teach.
What Is Guilford's Theory of Intelligence?
Guilford's Structure of Intellect theory is a three-part map of mental skills. Each proposed skill joins an operation, a type of content and a product. In plain terms, it asks what the mind does, what it works on and what form the result takes (Guilford, 1956).
This is a theory of how skills may be grouped. It is not a set of stages in child growth. It is not a proven profile of fixed strengths or needs. The wider guide to theories of intelligence sets it beside other accounts.
Guilford did not accept the view that one broad factor, known as g, could sum up the shape of the mind. He looked for many small factors instead. He then placed them in a grid. This made his model broad and clear, but it also made the test of that grid vital.
A cell in the grid is a proposed skill. It is not proof that such a skill exists on its own. Guilford and his team made tests for many cells. Critics then asked if the test method had been steered towards the shape he sought.
How the Model Developed
Guilford's model grew over almost four decades. His 1950 speech dealt with creative thought, not a full cube. The 1956 paper set out a new system that was still being built.
The well-known 120-cell form came in 1967. The last change came in 1988.
In his 1950 speech to the American Psychological Association, Guilford said that the field had paid too little heed to creativity. He also said that standard IQ tests did not show all forms of creative skill (Guilford, 1950). The speech helped make creativity a major field of study. It did not invent all thought about open or varied ideas.
The 1956 paper gave a new map of the adult mind. It used the three main parts of the later model. Yet it was still a work in progress.
It drew on about forty factors then found in past work (Guilford, 1956). It did not prove a cube of 120 distinct skills.
Guilford gave the mature cube in The Nature of Human Intelligence (Guilford, 1967). It had five operations, four forms of content and six products. Their possible joins gave 120 cells.
The neat count made the model easy to show. The count itself was not test proof.
He then split figural content into visual and sound-based content. This gave five content types and 150 cells. His last paper split memory into recording and retention.
The final grid had six operations, five contents and six products. That made 180 possible cells (Guilford, 1988).
The Three Dimensions
The model has three dimensions: operations, contents and products. Operations name what the mind does. Contents name the kind of data used. Products name the form of the result, and each full cell uses one term from each list.
Operations are acts of thought. The 1967 list had cognition, memory, divergent production, convergent production and evaluation. In 1988, memory became two acts: recording and retention. This gave six operations (Guilford, 1967; Guilford, 1988).
Cognition means finding or grasping data. Divergent production yields a range of ideas from what is given. Convergent production seeks an answer that fits the limits of the task.
Evaluation checks data against a rule or need. In Guilford's grid, convergence and evaluation are not the same act.
Contents are the kinds of data used in thought. The 1967 list had figural, symbolic, semantic and behavioural content. Guilford later split figural content into visual and sound-based content. The last model thus had five kinds.
Symbolic content includes signs such as letters and digits. Semantic content deals with meaning. Behavioural content deals with signs found in our acts with other people.
These are parts of Guilford's theory. They are not proof of visual, verbal or social learner types.
Products are the forms that data can take. The six types are units, classes, relations, systems, transformations and implications. A unit is one item.
A class is a group with a shared trait. A relation links items, while a system joins many links.
A transformation is a change in data. An implication is a likely result or claim that follows. In the model, a cell joins one operation, one content and one product.
The label can help show the claim. It does not prove that the cell is a separate skill.
From 120 to 180 Abilities
The totals changed when Guilford split parts of the grid. They are not stages of growth, groups of people or test scores. The order is 120, 150 and then 180. On a phone, swipe across the table to read each column.
On a smaller screen, swipe across to read all columns.
| Version | Operations | Contents | Products | Count and meaning |
|---|---|---|---|---|
| 1967 mature model | Five: cognition, memory, divergent production, convergent production and evaluation | Four: figural, symbolic, semantic and behavioural | Six: units, classes, relations, systems, transformations and implications | 5 × 4 × 6 = 120 proposed cells. This is the cube in Guilford's 1967 book. |
| Intermediate expansion | Five, with memory still in one part | Five, after figural content was split into visual and auditory | The same six products | 5 × 5 × 6 = 150 proposed cells. This rise came from the content split, not the memory split. |
| 1988 final revision | Six, after memory was split into recording and retention | Five: visual, auditory, symbolic, semantic and behavioural | The same six products | 6 × 5 × 6 = 180 proposed cells. Guilford set out this last form in 1988. |
The order matters. The first rise came from the split in figural content. The second came from the split in memory.
If those changes are reversed, the history is wrong. The error also hides why Guilford changed the model.
Each count shows all joins that the grid permits. It does not show how many cells had sound test support. Guilford claimed support for many small skills. The key issue was whether his method could test the whole shape in a fair way.
Divergent Production and Creativity
Divergent production is the best-known part of Guilford's work. It means making a range of ideas from the facts or prompt at hand. Convergent production seeks a fit within set limits.
Evaluation checks the result. Guilford kept those last two acts apart.
The 1950 speech said that IQ tests did not show all forms of creative skill (Guilford, 1950). Later work linked divergent production with fluency, flexibility, originality and elaboration. These terms had a large effect on research. They are not four quick marks that a teacher can score by eye.
Fluency is the count of apt ideas. Flexibility is the spread across groups of ideas. Originality deals with ideas that are both rare and apt.
Elaboration is the detail used to build an idea. Scores can change with the task, the sample, the time allowed and the scoring rule.
Runco and Acar call divergent thought a sign of creative potential (Runco & Acar, 2012). It is not the same as creative work. A high test score does not ensure that a person will make work of worth. Such work also needs skill, care, time, choice and a field in which it can count.
Scores for rare ideas pose a further risk. A person who gives more answers has more chances to give a rare one. What counts as rare also rests on the sample. A rare answer may still be poor or off the point (Silvia et al., 2008).
This does not make open tasks of no use. It sets a firm limit on what the result can show. An open prompt is one kind of task.
A sound claim about a person's creative potential needs a sound test. Wider use of such tasks belongs in the guide to creativity in education.
Guilford's Structure of Intellect Study Notes
This one-page note shows the three dimensions. It gives the right 120 to 150 to 180 order. It also keeps the test limits clear.
Read the accessible study-note transcript
Direct answer: Guilford's Structure of Intellect theory groups proposed mental skills by operation, content and product. It is a past model, not a proven learner profile or way to teach.
Operations: the 1967 model had cognition, memory, divergent production, convergent production and evaluation. In 1988, memory was split into recording and retention. This gave six operations. Convergent production and evaluation stay apart.
Contents: the 1967 model had figural, symbolic, semantic and behavioural content. The split of figural content into visual and sound-based forms gave five contents.
Products: units, classes, relations, systems, transformations and implications name the proposed form of the result.
Correct order: five operations by four contents by six products gave 120 cells in 1967. The visual and sound-based content split gave 150. The later memory split gave six by five by six, or 180, in 1988.
Divergent production: Guilford made the open flow of varied ideas a key part of modern work on creative thought. Fluency, flexibility, originality and elaboration became well-known signs. Yet such scores show potential, not creativity itself.
Evidence limit: Horn and Knapp showed that the method used to back SOI could make random theories look sound. Carroll's large review found a layered shape with broad and narrow skills plus a general factor. This weakens the full cube but does not erase its place in history.
Use in school: an open prompt may help vary a task. The grid should not label learners, find gaps, dictate support or imply that work on one cell builds a distinct form of intelligence.
Evidence and Factor-Analytic Criticism
The full grid rests on factor analysis. This set of methods looks for shared trends in test scores. The main test is simple to state: can a fair method find the proposed shape without being told to seek that shape? Critics said Guilford's method failed this test.
Limitations and Critiques
Horn and Knapp looked at studies used to back SOI. They challenged the use of target, or Procrustean, factor rotation. With this method, an analyst turns the factors towards a form set in advance. They showed that it could make random theories look as if the data backed them (Horn & Knapp, 1973).
This is a grave flaw in the case for the full cube. It does not mean that each task had no value. It does not mean that all small skills named by Guilford were false. It means that this set of results could not give a fair test of the whole model.
One paper did not settle each part of the field for all time. The case changed through many tests, new methods and new models. The sound claim is more narrow. Support for the full SOI grid was weak, and a main method used to confirm it was open to a strong charge.
This point keeps two facts apart. The cube is a clear map of Guilford's claims. It is not proof of those claims. At the same time, later work could still study small skills such as fluency without taking on the whole cube.
Relationship to Modern Ability Models
Later work found a layered shape for mental skills. It did not find a flat grid of equal cells. Carroll's large review had three levels. Small skills sat under broad groups, with a general factor at the top (Carroll, 1993).
Carroll used hundreds of past sets of test data and one common method. His broad groups covered fluid and learned skill, memory, sight, sound, recall and speed. The page on fluid intelligence explains one of these later strands. It is not just a new name for an SOI cell.
The Cattell-Horn-Carroll family drew a range of past work into one broad frame. McGrew set out this link and its growth over time (McGrew, 2009). Cattell-Horn-Carroll theory now has a major role in tests of mental skill. Yet it is still a live family of models, not the last word on the mind.
Other plural theories do not share Guilford's grid. Sternberg's triarchic theory uses a different set of claims. So does Gardner's multiple intelligences. All three cast doubt on a bare score, but this does not make their proof or terms the same.
Guilford's place is thus mixed. He made the field ask about a much wider range of skills. He gave creative thought a new and clear role. But the full grid did not become the main base for modern tests of intelligence.
What Education Can and Cannot Take
Schools can take one small question from Guilford: what sort of answer does this task seek? Some tasks have a tight goal. Some ask for a range of sound ideas.
Both may be useful. This is a way to describe a task, not proof of an SOI method.
A class may first list several views of a text. It may then check each view against the text. This shows a shift from open ideas to a test with rules. It does not prove that such a shift builds two types of intelligence or raises marks.
The grid cannot tell a school which cell fits a learner. It cannot prove that a learner has a visual or semantic style. It cannot find a need from one piece of class work. Nor can it show that work on one cell will build a distinct skill in the mind.
Teachers should not mark originality by picking the rarest answer. Rare does not mean apt. A learner who gives more ideas has more chances to give an odd one (Silvia et al., 2008). A formal test needs sound rules, fair use and skilled judgement.
Guilford still has worth as a part of the history of ideas. His terms can also help a scholar compare open production, tight production and judgement. The guide to critical thinking deals with sound judgement in more depth. The Learning Theories collection sets this model next to other views and their limits.
Frequently Asked Questions
The cell count depends on the version of the model. Its three dimensions are operations, contents and products. Divergent thought is not the same as creative work. The full grid has weak test support, and its use in schools must stay descriptive rather than diagnostic.
How many abilities are in Guilford's theory?
The 1967 model had 120 proposed cells. The split of figural content into visual and auditory forms gave 150. The split of memory into recording and retention then gave 180 in 1988 (Guilford, 1967; Guilford, 1988).
What are the three dimensions of the Structure of Intellect model?
The dimensions are operations, contents and products. Operations tell us what the mind does. Contents tell us what kind of data it uses.
Products tell us what form the result takes. One term from each set gives one proposed cell.
Is divergent thinking the same as creativity?
No. Tests of divergent thought can show parts of creative potential. They do not ensure creative work (Runco & Acar, 2012).
Creative work must also be apt and useful in its field. It draws on skill, care, choice and the chance to build an idea.
Is Guilford's Structure of Intellect model supported by evidence?
The case for the full grid is weak. Horn and Knapp showed that its test method could make random theories look sound (Horn & Knapp, 1973). Later large reviews found a layered shape instead. Some small skills and Guilford's work on creative thought still had a life of their own.
Can teachers use the SOI model in the classroom?
It can help scholars talk about the history of ideas. It may also prompt a broad look at task demands. It should not label learners, find gaps, dictate support or claim that work on one cell builds a distinct intelligence.
References
- Carroll, J. B. (1993). Human Cognitive Abilities: A Survey of Factor-Analytic Studies. Cambridge University Press.
- Guilford, J. P. (1950). Creativity. American Psychologist, 5(9), 444-454.
- Guilford, J. P. (1956). The Structure of Intellect. Psychological Bulletin, 53(4), 267-293.
- Guilford, J. P. (1967). The Nature of Human Intelligence. McGraw-Hill.
- Guilford, J. P. (1988). Some Changes in the Structure-of-Intellect Model. Educational and Psychological Measurement, 48(1), 1-4.
- Horn, J. L., & Knapp, J. R. (1973). On the Subjective Character of the Empirical Base of Guilford's Structure-of-Intellect Model. Psychological Bulletin, 80(1), 33-43.
- McGrew, K. S. (2009). CHC Theory and the Human Cognitive Abilities Project. Intelligence, 37(1), 1-10.
- Runco, M. A., & Acar, S. (2012). Divergent Thinking as an Indicator of Creative Potential. Creativity Research Journal, 24(1), 66-75.
- Silvia, P. J., Winterstein, B. P., Willse, J. T., Barona, C. M., Cram, J. T., Hess, K. I., Martinez, J. L., & Richard, C. A. (2008). Assessing Creativity with Divergent Thinking Tasks: Exploring the Reliability and Validity of New Subjective Scoring Methods. Psychology of Aesthetics, Creativity, and the Arts, 2(2), 68-85.
