本研究的目的在於運用驗證性因素分析測試產前依附量表(Prenatal Attachment Inventory, PAI)的建構效度。PAI乃由Muller(1989, 1993)所發展的工具，共有21題的四分量表。研究樣本為來自於產科門診、住院及產前生產教育班的349位第三孕期婦女。以驗證性因素分析發現：PAI 的所有21個項目的因素負荷量皆達統計上顯著水準、沒有負的誤差變異、因素間的相關性沒有大於1、沒有極大的誤差變異；而且個別項目信度介於.2至.5之間。另外，相關係數矩陣顯示：所有的21個項目皆達統計上的關聯性(p<.01)，共解釋79%的變異量。評估量表的整體模式適合度，發現假設的模式在經過修正指標修正後，PAI量表便具有相當好的合適度的測量模式，÷2/df = 2.12、根號平均平方餘差(standardized root mean square residual)為.05、最佳適合度指標(goodness-of-fit index)為.90、臨界N值(critical N)為204.5。本研究結果證明PAI為一個具有效度的量表，可以供研究者及臨床專家用來測量及評估母胎依附關係。

The pur pose of this study was to test the con struct valid ity of the Pre na tal Attach ment Inven tory (PAI) by using Con fir ma tory Fac tor Anal y sis (CFA). The PAI was devel oped by Mul ler (1989, 1993), and con sists of 21 Likert-type items arranged on a 4-point response set rang ing from “almost always” to “almost never.” A sam ple of 349 preg nant women in their third tri mes ter was recruited from obstet ri cal out pa tient and inpa tient cen ters, includ ing high-risk and low-risk preg nancy units, and child birth edu ca tion classes. Con fir ma tory fac tor anal y sis of the instru ment yielded the result that all items have a sig nif i cant fac tor load ing. There were no neg a tive error vari ances, no fac tor cor re la tions greater than one, and no extremely large error vari ance. The indi ca tor reliabilities ranged from .2 to .5. More over, the cor re la tion matrix indi cated that all PAI items were sig nif i cantly related to one another at p <.01. These 21 items explained 79% of vari ance in mater nal fetal attach ment (MFA). The over all mea sure ment model fit indi ces showed that the hypoth e sized model fit the data well after two mod i fi ca tions (c2/df = 2.12, stan dard ized root mean square resid ual = .05, good ness-of-fit index = .90, crit i cal N = 204.5). The find ings indi cate that Mul ler’s PAI is a valid instru ment for research with preg nant women. Researchers can use this instru ment for future research stud ies in the health care of preg nant women. Nursing edu ca tors and cli ni cians can use this instru ment to assess expec tant moth ers’ level of MFA.