因逕分析的概念與方法

Concept and Method of Path Analysis

10.7081/NR.200002.0124

王秀紅(Hsiu-Hung Wang)

因逕分析 ； 因果模式 ； 因逕係數 ； path analysis ； causal modeling ； path coefficients

護理研究

8卷1期（2000 / 02 / 01）

124 - 133

繁體中文

雖然有許多合適的技巧來發展及檢驗理論模式，對於以實務爲基礎的護理學科而言，因逕分析是用來理論與實證資料相互印證相當實用的一種方法。運用適宜，因果模式可用來補足理論缺乏實據驗證或實證研究缺乏理論基礎的缺憾。因逕分析常被視爲是檢驗因果模式之精確性與代表性之一種方法。因逕分析應用於護理研究最早可追溯至1980年代，直到目前爲止，已經有許多護理研究者利用因逕分析來處理資料或檢驗理論模式。因逕分析在護理研究的應用有助於辨識變項之間錯綜複雜的關係網路。當護理研究者逐漸投入複雜的理論模式發展階段，利用較高級的資料分析技巧檢驗變項之間錯綜複雜的關係，可更有效地解釋或預測人類行爲或反應的特殊現象，並發展適當的護理照護模式。本文將針對因逕分析的假說、因逕分析的步驟（包括：發展一植基於理論的模式、建構一因果相關的因逕圖、決定樣本數的大小、選擇適當的矩陣型式輸入電腦、模式的檢驗與修飾、模式適宜度的評估、結果的解釋）、因逕分析的優點與限制、以及因逕分析在護理的應用等加以說明與討論。

Although many legitimate techniques can be used to develop and test theoretical models, for a practice- oriented nursing discipline, path analysis is a very useful way to examine the on-going interplay between theoretical specification and empirical verification. If it is appropriately used, causal modeling can complement both the inadequacies of theories without empirical verification and atheoretical empiricism. Path analysis is considered a precise and representative method of causal modeling for theory testing and generation. The first application of path analysis in nursing research can be traced back to 1980. Until recently, many nurse researchers have begun to use path analysis for data analysis or theoretical model testing. In nursing, path analysis is very useful in distinguishing the complex relationships among variables. As nurse researchers make efforts to develop complex theoretical models, using advanced strategies for data analysis to test the complex relationships among variables can help explain and predict the phenomena of human behaviors and responses and then develop appropriate nursing care models. This article focuses upon presenting and discussing the assumptions of path analysis, the steps of path analysis (including the development of a theory-based model, construction of a path diagram, determination of the sample size, selection of an appropriate matrix for input, model testing and modification, assessment of model fit, and explanation of the results), the advantages and limitations of path analysis, and the applications of path analysis in nursing.

1. Beaton, R. D., Murphy, S. A., Pike, K. C.,Corneil, W.(1997).Social support and network conflict in firefighters and paramedics.Western Journal of Nursing Research,19(3),297-313.
2. Biddle, B. J.,Marlin, M. M.(1987).Causality, confirmation, credulity, and structural equation modeling.Child Development,58(1),4-17.
3. Bollen, K. A.(1989).Structural equations with latent variables.New York:John Wiley & Sons.
4. Boyd, C. J.,Frey, M.,Aaronson, L. S.(1988).Structural equation models and nursing research: Part I.Nursing Research,37(4),249-252.
5. Byrne, B. M.(1994).Structural equation modeling with EQS and EQS I Windows: Basic concepts, applications, and programming.Thousand Oaks, CA:Sage.
6. Conn, V. S.(1998).Older adults and exercise: Path analysis of self-efficacy related constructs.Nursing Research,47(3),108-189.
7. Ferketich, S. A.,Verran, J. A.(1990).Advancing nursing science through research.Newbury Park, CA:Sage.
8. Hair, J. F.,Anderson, R. E.,Tatham, R. L.,Black, W. C.(1992).Multivariate data analysis.New York:Macmillan.
9. Hickey, M. L.(1993).Statistical methods for health care research.Philadelphia:I. B. Lippincott.
10. Joreskog, K.,Sorbom, D.(1996).LISREL 8: Users reference guide.Chicago:Scientific Software International.
11. Kurlowicz, L. H.(1998).Perceived self-efficacy, functional ability, and depressive symptoms in older elective surgery patients.Nursing Research,47(4),219-226.
12. Loehlin, J. C.(1987).Latent variable models: An introduction to factor, path, and structural analysis.Hillsdale, NJ:Lawrence Erlbaum Associates.
13. Mason-Hawkes, J.,Holm, K.(1989).Causal modeling: A comparison of path analysis and LISREL.Nursing Research,38(5),312-314.
14. Munro, B. H.(1980).Dropouts from nursing education: Path analysis of a national sample.Nursing Research,29(6),371-377.
15. Page, E. B.(1993).Statistical methods for health care research.Philadelphia:J. B. Lippincott.
16. Phillips, L. R.(1990).Causal modeling as a relevant approach to gerontological nursing.Journal of Gerontological Nursing,16(9),20-24.
17. Preski, S.,Walker, L. O.(1997).Contributions of maternal identity and lifestyle to young children`s adjustment.Research Nursing Health,20(2),107-117.
18. Stevens, J.(1996).Applied multivariate statistics for the social sciences.Mahwah, NJ:Lawrence Eribaum Associates.
19. Tanaka, J. S.(1987).How big is big enough: Sample size and goodness of fit in structural equation models with latent variables.Child Development,58(1),135-146.
20. Thorndike, R.(1978).Correlational procedures for research.New York:Gardner Press.
21. Wang, H. H.(1998).A model of self-care and wellbeing of rural elderly women in Taiwan.Unpublished doctoral dissertation, University of Texas at Austin.
22. Wang, H. H.,Lee, I.(1999).A path analysisof self-care of elderly women in a rural area of southern Taiwan.Kaohsiung Journal of Medical Sciences,15(2),94-103.
23. Weaver, T. B.,Richmond, T. S.,Narsavage, G. L.(1997).An explanatory model of functional status in chronic obstructive pulmonary disease.Nursing Research,46(1),26-31.
24. Winslow, B. W.(1997).Effects of formal supports on stress outcomes in family caregivers of Alzheirmer`s patients.Research in Nursing and Health,20(1),27-37.
25. Woods, N. F.(1988).Nursing research: Theory and practice.St. Louis:C. V. Mosby.
26. Wright, S.(1921).Journal of Agricultural Research.
27. 黃璉華(1992)。老人生活滿意度相關因素之因徑分析研究。護理雜誌，39(4)，37-47。
28. 蕭淑貞、鄭麗香、蔡世澤、陳月枝、林瑞祥、戴東原、梁靜祝(1995)。以因徑分析法檢定影響胰島素依賴型糖尿病控制的因素。公共衛生，22(1)，9-15。