青少年希望感量表之建構

The Development of Hope Scale for Teens

本研究的主要目的在建立一份具備信效度的青少年希望感量表。研究者進行國內外相關文獻探討，並參考Snyder、Harris、Anderson 及Holleran 等人於1991年所發展的「成人的希望感傾向量表(Adult Dispositional Hope Scale)」以及Sympson(1999)所發展出來之「特定領域的希望感量表(Domain Specific HopeScale)」之題目設計進行問題編擬。蒐集全國在學國中及高中青少年共1052 份有效問卷，以LISREL 8.72 統計軟體進行驗證性因素分析，首先為估計方法檢測，並以競爭模式方式針對觀察資料與模式彼此之間適配情形進行最適配模式之選取，本研究將採用之競爭模式分別為虛無模式、單因素模式、多因素直交模式、多因素斜交模式以及多階層模式。經過競爭模式指標後採用多因素斜交模式為最簡效模式，針對此模式進行內在結構適配評鑑，結果顯示觀察變項與潛在變項具備良好的信度與尚可聚合效度、區別效度。為檢驗此多因素斜交模式的穩定度，將整體樣本隨機區分為兩個樣本，分別為檢定樣本(N=526)及效度樣本(N=526)進行複核效化的評鑑，結果顯示嚴謹策略下，五因素斜交模式之青少年希望感量表具有穩定性以及預測力。

The purpose of this study was to develop the reliability and validity of Hope Scale for Teens. This scale was developed by the authors， based on the reference: Adult Dispositional Hope Scale and Domain Specific Hope Scale.LISREL 8.72 was used for Confirmatory Factor Analysis， competition models included null model， one-factor model， uncorrelated factors model， correlated factors model and hierarchical model. Confirmatory Factor Analysis showed that there was an acceptable goodness of fit between the scale correlated factors model， and this research with good reliability and validity. Using a random split， the sample(N=1052) was divided into a calibration sample(N= 526) and a validation sample(N= 526).The results of cross-validation suggested that the five-factor model was supported inboth samples under the tight replication strategy.