Abstract
Brainerd's (1973) cardination test is supposed to assess children's understanding of one‐to‐one correspondence independently of their counting ability. Children's errors in this test could result from difficulty in executing a pairing strategy rather than defective understanding. This was tested by comparing performance on the cardination test displays with performance on such displays with guidelines added to make pairing easier. Both 5‐year‐olds and 7‐year‐olds took part. Guidelines helped both age groups; the 7‐year‐olds more than the 5‐year‐olds. There were very marked differences between the age groups.