Raven’s Progressive Matrices (often referred to simply as Raven's Matrices or RPM) are multiple choice tests of abstract reasoning, originally developed by Dr John C. Raven in 1936. In each test item, a person is asked to analyze a geometric pattern and identify the missing piece. After a few very easy inital patterns, the subsequent geometric patterns are 3x3 matrices (giving the test its name).
According to their author—and corroborated by decades of research—Raven’s Progressive Matrices and Vocabulary tests measure the two main components of general intelligence (originally identified by Charles Spearman): the ability to think clearly and make sense of complexity, which is known as educative ability, and the ability to store and reproduce information, known as reproductive ability.
The RPM is a completely non-linguistic and non-mathematical test: there are no letters, no numbers, just patterns of geometric shapes. For this reason it can be used with non-English speakers. The RPM has been used in many cultures and languages around the world, and research has shown that it is culturally neutral. Adequate standardization, ease of use (without written or complex instructions), and minimal cost per person tested are the main reasons for its widespread international use in many countries around the world.
Some of the most fundamental research in cognitive psychology has been carried out with the RPM. The tests have been shown to work—scale—measure the same thing —in a vast variety of cultural groups. Two remarkable, and relatively recent, findings are that, on the one hand, the actual scores obtained by people living in most countries with a tradition of literacy —from China, Russia, and India through Europe to Kuwait— are very similar at any point in time. On the other hand, in all countries, the scores have increased dramatically over time such that 50% of our grandparents would be assigned to special education classes if they were judged against today's norms. Yet none of the common explanations—access to television, changes in education, changes in family size etc.—hold up. The explanation seems to have more in common with those put forward to explain the parallel increase in life expectancy which has doubled over the same period of time
John Carlyle Raven first published his Progressive Matrices in the United Kingdom in 1938. His three sons established Scotland-based test publisher J C Raven Ltd. in 1972. In 2004, Harcourt Assessment, Inc. a division of Harcourt Education acquired J C Raven Ltd.
The Progressive Matrices:
The Standard RPM is a series of 60 matrices, in 5 sets of 12 each (A1–A12, B1–B12 … to E1–E12). They begin with very easy patterns, in which all is needed is the ability to comprehend the concept of a missing piece, and the ability to see. They get progressively more complex, their logic increasingly difficult to discern. Here are a few examples:
|The A & B matrices are quite simple:|
|The D & E matrices are quite challenging:|
There are a few online versions of the RPM. Test yourself! Try www.egopont.com/iq/index.php, www.iqtest.dk/main.swf, or www.clipsite.com.ar/HOME/Salud/Test/Raven/ (en español). The test at www.similarminds.com/intdoor.html is similar but not as close to the RPM as the others.
Scoring the Ravens:
After a person has completed the Ravens test, their answers are corrected and the total number of correct responses determined. Then this “raw score” is checked against a scoring matrix table to determine the nationally-normed percentile score to which their raw score correlates. These percentile scores are based on the age of the tester; the scoring matrix groups testers into 6-month groups and is validated for children ages 6½ to 16½.
There is also a secondary scoring process, in which the number of correct resposnes per set of 12 matrices is compared to nationally-normed expectations for each set, to determine consistency of answers. The Ravens Manual1 states that if the discrepancy between a tester's score and expected score for any one section is greater than +2 or -2, then the overall score is less reliable “as a consistent estimate of general capacity for intellectual activity. [However,] for general purposes the total score appears to be relatively valid even when discrepancies of more than two points occur.” It’s hard to interpret this seeming contradiction, and we in the GATE Services Office in PVUSD do not calcualte these per-set discrepancies, and instead just use the total raw score to determine percentile score.