Abstract: Objective To reassess the reliability of the School-age Children′s Eating Behavior Scale developed by the research group by the generalizability theory,and then to revise the original scale. Methods Totally 700 parents of children aged 6 to 12 years were randomly selected from two primary schools in urban and suburb of Xi′an from March to June 2018,and were asked to filled out School-age Children′s Eating Behavior Scale.Then the scale was evaluated and revised based on generalizability theory. Results The generalized coefficients of the 8 dimensions were 0.74-0.85,and the reliability indexes were 0.73-0.84,which were greater than 0.70 and reaching the good level.The performance of dimensions of ill-balanced eating,food responsiveness,satiety responsiveness,craving for junk food were relatively good,while those of worse eating habit,external factor eating,dietary restriction,food preference were relatively worse.According to experts and GT analysis results,the final revision scale,including 46 items and 8 dimensions,had good reliability and validity. Conclusions The School-age Children′s Eating Behavior Scale revised and improved based on the generalizability theory,is prove to be of good test accuracy and reliability.Therefore,it could be used to measure and evaluate the eating behaviors of school-age children in China.

Key words: school-agechildren, EatingBehaviorScale, generalizabilitytheory

摘要： 目的 运用概化理论对本研究组前期编制的“学龄期儿童饮食行为量表”进行再次的信度评估,并进行进一步的修订。方法 2018年3-6月在陕西省西安城、郊2所小学随机抽取的700名6～12岁儿童家长进行“学龄期儿童饮食行为量表”调查,采用概化理论进行量表评估和修订。结果 8个维度概化系数介于0.74～0.85,可靠性指数介于0.73～0.84,均大于0.7,达到了理想水平。挑食、食物响应、过饱响应、垃圾食品渴望维度条目质量相对较好,而不良进食习惯、外因性进食、限制饮食、食物喜好维度4个维度质量相对逊色,按照专家讨论分析及GT理论结果的支持,最终修订形成46个条目、8个维度、信效度更为良好的“学龄期儿童饮食行为量表”。结论 经过概化理论最终修订完善的“学龄期儿童饮食行为量表”具有较好的测验精度和信度,可用于我国学龄期儿童饮食行为的测量及评估。

关键词: 学龄期儿童, 饮食行为量表, 概化理论