本研究目的在於編製適合我國老年人之寂寞量表。研究首先以個別及焦點團體訪談方式，訪談了16位老年人，並將訪談內容分析整理，以及參考相關文獻及量表，進而形成量表題目。研究接著以187位老年人，進行量表之項目分析、探索性因素分析、因素間的相關分析、效標關聯效度及內部一致性信度分析，結果可區分為「情緒性寂寞」與「社會性寂寞」兩個分量表，共計10題，全量表α係數為.86，分量表α係數為.87與.62。本研究另以242位老年人，進行驗證性因素分析及組合信度等考驗，結果發現模式適配度良好，組合信度亦可接受，因此，老年人寂寞量表具有可接受的信效度。最後本研究根據研究結果進行討論，並對後續研究提出建議。

The purpose of this study was to development the Elderly Loneliness Scale（ELS）and examined its reliability and validity. The researcher first selected 16 elderly for personal and focus group interviews, then analyzed the qualitative data and reference related scales to create the items of the ELS. The study used 187 participants to conduct item analysis, exploratory factor analysis, factor-to factor correlation, criterion-related validity and reliability analysis. ELS consists of 10 items, and can be divided into two subscales, namely emotional loneliness and social loneliness, and Cronbach，s α efficient of total scale and two subscales were . 86、. 87 and . 62. The researcher then used the other 242 participants to conduct confirmatory factor analysis and composite reliability. The results reveal that the ELS has good goodness-of-fit, and two subscales have acceptable composite reliability. Finally, the implications and limitations of this present study were discussed.