兒童內化行為問題測量工具的建立及其信效度研究

The Development of Children Internalizing Problem Behavior Scale:A Reliability and Validity Study

本研究旨於編製一份適用於評量兒童內化行為問題的量表，並分析其信效度。兒童內化行為問題包括：憂鬱、社交焦慮、社交孤寂等三個構念，共有 20 題，為三點量表。採用「兒童及青少年行為之長期發展研究」第一年資料，分成國小一年級與四年級兩組對象進行信效度分析。結果顯示，內化行為問題分量表及總量表 值為 .66～.85。分析後發現，一年級與四年級在刪除「一個人會覺得寂寞」此題項後，模型在一年級有可接受的整體配適度；四年級則在估計「心情不好」與「覺得想哭」兩題項間之相關誤差後，呈現可接受的整體配適度。結果亦發現，一年級與四年級兩模型具聚斂效度與區別效度，且有可接受的標準因素負荷值、指標信度與綜合信度估計值。結果支持兒童能自我評量其內化行為問題，此量表有良好的信度與效度。

The main goals of this study were:  to develop an internalizing problem behavior scale for children and  to evaluate the reliability and validity of this scale. Originally，there were 20 items (3-point Likert type) in this scale to measure three aspects of inter-nalizing problem behaviors: depression， social loneliness， and social anxiety. Two co-horts of data (the first and the fourth graders) collected in 2001 from the ＂Child and Ad-olescent Behaviors in Long-Term Evolution (CABLE)＂ project were used to validate this scale.For the first and the fourth graders， the internal consistency coefficients of the whole scale and three subscales ranged from .66 to .85. A series of confirmatory factor analy-ses (CFA) were performed to evaluate the fit of the 3 construct-driven factors and the validity/reliability of the items used to measure each factor. For both graders， the item “You feel lonely when you are alone” was foundtohavecross-loadings.Afterdeleting this item， the 3-factor model with 19 items provided acceptable fit for the first graders.For the fourth graders， one correlated error needed to be estimated in the 3-factor， 19-item CFA model to obtain a good fit.Generally speaking， the internalizing problem behavior scale exhibited acceptable values for indicator reliability， composite reliability， convergent validity and discrimi-nant validity. The self-report inventory of internalizing behavior problems seems to be a useful and valid tool applied to children.