從事志願役態度量表之發展

Development of an Attitude Scale about Persons in Volunteer Military Servicemen

蔡佩圜(Tsai, Pei-Huan)

《國防管理學報》 41卷1期 (2020/05) Pp. 53-76

摘要

本研究之目的是在發展從事志願役態度量表，全量表試題（14題），可分為「激勵因素」與「阻礙因素」2個向度，在預試階段（樣本數為300）藉由Rasch模式之內、外適合度檢驗受試者與本文所發展的量表之適切性，結果顯示量表所有試題平均的Aiken內容效度達.767，Rasch模式適合度檢驗均在理想範圍內。在正式施測階段以隨機取樣方式選取其中一組（樣本數為755）首以探索性因素分析，進一步將量表試題刪減為13題，結果顯示量表涵蓋2個向度符合本文之研究設計，其分量表信度分別為.882、.851，餘量表所有試題（13題）平均的Aiken內容效度達.760；另一組（樣本數為756）次以驗證性因素分析方式檢驗雙因素模式之適配度，藉由比較競爭模式之適配情形進一步將本量表精簡成8題試題，餘量表所有試題（8題）平均的Aiken內容效度達.849，具收斂效度與區別效度，分測量之組合信度分別為.900、.774，能滿足跨軍職身分群組測量不變性檢驗。因此，本研究所發展之精簡版量表試題（8題）具有良好的信度與效度，可作為探討影響從事志願役態度之評量工具。

關鍵字

內適合度 ； 外適合度 ； 收斂效度 ； 區別效度

並列摘要

The purpose of this study is to develop an attitude scale on soldiers of volunteer military service and to check the Rasch model fit of the two subscales in the pretest phase(N=300). The results show that the validity of Aiken's content of the average of all the scale items is .767. Infit mean-square(InMnsq) and outfit mean-square(OutMnsq) of all items verify fittingness. The formal testing phase, a group(N=755)analyzed by EFA, is reduced to 13. The results show that the scale covers two dimensions, the subscale reliability is respectively .882 and .851, the validity of Aiken's content of the average is .760. In addition, the other group(N=756) checks the two-factor pattern fit with CFA, and the scale is reduced to 8 through the comparative competition model. The validity of Aiken's content of the average is .849, with a convergence validity and a discriminant validity. The Composite Reliability(CR) of the subscale is respectively .900 and .774, which can content with the test-invariance of cross-military identity groups. In brief, the scale developed by this study could be regard as a good reliability and validity material, and it can be used as a measurement tool to analyze the attitude of people in volunteer military servicemen.

並列關鍵字

Convergent Validity ； Discriminant Validity ； Infit Mean-square ； Outfit Mean-square