Research on education often faces the problem of measurement of the op
portunity to learn. This is also the case with mathematics education.
Researchers may choose either a retrospective measurement approach (e.
g., a questionnaire) or an immediate recording approach (e.g., logbook
s). In the present study, the accuracy, reliability, and predictive va
lidity of a logbook procedure is assessed. The results show that the u
se of logbooks leads to reliable measurement of opportunity to learn v
ariables. The variables measured by logbooks also appear to be good pr
edictors of learning progress.