The effect on concept mapping of instructions to make linking labels was ex
amined using the "atom" unit of senior high school chemistry. 64 randomly s
elected students were divided into two groups. Both groups were given 14 ch
emistry concepts and told to make concept maps, but only one group was inst
ructed to make linking labels. Multiple-choice tests with 12 equivalent ite
ms were given to the students before and after the mapping. After the prete
st, students in each group were reassigned into two groups according to the
ir scores on a test administered before the pretest. A 2-factorial analysis
of covariance was carried out to find the interaction between differences
due to the instruction to make linking labels and scores on this test. In t
he uninstructed condition, the mean ANCOVA score and scores on this test we
re consistent, whereas in the condition instructed to make linking labels,
the students with lower scores on this test improved as much as the student
s with higher scores.