督導關係量表之編製與信、效度建構

Development and Validation of a Supervisory Relationship Scale

本研究旨在編製一個適用於受督者之「督導關係量表」，以協助督導者或受訓中的督導者，檢視其與受督者知覺的督導關係發展之狀況和得失所在，作為改進和提昇督導關係之參考依據，並可提供未來研究用以探究和理解督導關係和其他各種督導議題的關聯。本研究主要根據先前研究成果，並參酌實務經驗、以及學者專家論述，發展量表架構與題項，共78題。再針對量表初編題項進行專家效度檢核後，依建議修正，並刪除CVI小於.80之題項，共得預試題項52題。預試時針對不同背景相關心理專業受督者抽樣，得198份有效樣本。預試後52題之臨界比（CR值）和內部一致性均達.001顯著水準，形成一個52題的量表。效度方面，經因素分析共得七個因素，在刪除僅有一個題項的第七個因素之後，其餘命名為「安全的督導環境與個人特質」、「值得信賴的督導與專業素養」、「促進認知發展的督導方式」、「形成性的回饋與角色典範」、「協助覺察與適性指導」、「督導者的自我揭露」等六個因素，形成一個51題的正式量表，總累積變異量為61.715%；與「督導關知覺量表」得分之積差相關係數γ=.871（p

The purpose of this study was to develop a valid and reliable measure of supervisory relationship from the perspective of supervisees. Qualitative data obtained from a previous study and information generated from a review of the literature were used to construct a 78 item self-report questionnaire measure. After an evaluation by six professionals who used a constructive model to analyze the content validity of the scale， 4 items were combined into 2 and 24 items with CVI lower than .80 were deleted， resulting in a 52-item original version of Supervisory Relationship Scale (SRS). The scale was administered to 198 counseling supervisees with diverse background characteristics on gender， training level， affiliated universities， institutions， and geographical areas. The principal components analysis identified seven components for the construct of SRS. Due to having only one single item， the seventh component was removed. Finally the SRS was comprised of six components and 51 items in total. The components were labeled: “Safe Supervisory Environment and Personal Characteristics”， “Trustworthy Supervision and Professional Traits”， “Supervision Intervention Facilitating Cognitive Development”， “Formative Feedback and Role Modeling”， “Helping Awareness and Situated Intervening”， and “Supervisor Self Disclosure” . The amount of variance in SRS explained by the six factors was 61.715%. The results of the analyses for the scale’s concurrent validity， internal consistency and split-half reliability were favorable. The SRS demonstrated good concurrent validity (γ = .871， p < .001)， high internal consistency (α = .978)， and good split-­half reliability (γxx = .959). On the basis of the findings of the study， the applications of this instrument for counselor and supervisor training and for future research are forwarded.