以老人居住安排的轉換是否與子女同住而言，可分為「由與子女同住轉為不與子女同住」和「由不與子女同住轉為與子女同住」二種類型。以往討論老人居住安排的轉換，大多以家庭價值觀解釋，強調感情因素，當老人有最大需要時，例如，健康惡化或從有偶變成無偶，更可能發生轉換為與子女同住。但現代化社會，經濟因素轉趨重要，用家庭價值觀解釋轉換，可能已有所不足，需要以交換理論補強。然而以往臺灣地區已完成之相關研究較少同時由這兩種觀點出發，以及缺乏適合分析之長期資料，致相關研究不足。因此，本文使用衛生署與密西根大學合作之「臺灣地區中老年身心社會生活狀況長期追蹤調查」縱斷面調查資料進行老人居住安排轉換的分析研究，觀察1989 年與1996 年前後7年之居住安排轉換行為，以檢視這兩種觀點對於居住安排轉換的解釋力。又，以往的相關研究大多以靜態變數為解釋變數，本文貢獻主要在於以動態的變數－相關變數前後期的變化－為解釋變數，檢視其對老人居住安排轉換的影響。結果發現：以家庭價值觀或以交換理論解釋老人居住安排的轉換，都獲得部分支持；但沒有足夠的證據支持健康惡化或從有偶變無偶是影響老人轉換為與子女同住的關鍵因素；相反地，與交換理論相關的重大經濟決策權的改變變得比較重要，造成更可能發生轉換為不與子女同住或與子女同住。隱含著家庭價值觀對老人居住安排轉換的影響逐漸式微，而交換理論的影響力則越來越強，可彌補以家庭價值觀解釋之不足。

The transition in living arrangements among the elderly based on whether to live with children is classified into "transition to live apart from children among those living with children at baseline" and "transition to live with children among those living apart from children at baseline". The value placed on family was often used to explain the influence on living arrangements transitions in previous literature. Yet economic factors have become important in modern society, therefore we need to use exchange theory to strengthen the explanation on transitions. However this argument was not often discussed in previous literature in Taiwan, and the related research was lacking, owing to the insufficiency of long-term data which may be used for analysis. This paper, therefore, attempts to compare these two explanations through analyzing transitions, and to examine the effects of these two viewpoints on transitions. Using a longitudinal data source, the Survey of Health and Living Status of the Elderly in Taiwan, produced by the Department of Health, R.O.C. (Taiwan) and the University of Michigan, a multiple logistic regression model is used to observe the behavior of transitions in the 7 years between 1989 and 1996. The contribution of this article mainly lies in taking the dynamic variables, the changes of related variables between baseline and follow-up, as independent variables to examine the influence on the transitions. The empirical results of this paper show that both family value and exchange theory cause the living arrangements transitions, but there is not enough evidence to prove that a senior's health deterioration or the loss of a spouse was the major determinant of the transition to live with children. On the contrary, transitions of living arrangements are more likely to occur when a senior changes his/her role as the economic decision-maker into the non economic decision-maker of the household or vice versa. That implies the influence of family values on transitions is decreasing, and the influence of exchange theory is increasing.