Fig. 3From: A school-based nutrition education program involving children and their guardians in Japan: facilitation of guardian-child communication and reduction of nutrition knowledge disparityDefinition of variables in linear mixed models. Differences in the definitions of three variables (Implementation of the program, Nutrition education group, Evaluation timing) are shown. “Implementation of the program (before vs after the program)” was a major independent variable in showing the effect of the nutrition education program. “Nutrition education group (First vs Second group)” was a covariate used to adjust the difference in knowledge at baseline between the two groups. “Evaluation timing (May vs October or February)” was also a covariate used to adjust the difference in knowledge in the time courseBack to article page