Bagozzi R P, Warshaw P R
Multivariate Behav Res. 1992 Oct 1;27(4):601-34. doi: 10.1207/s15327906mbr2704_6.
The nature of the attitude-behavior relation was investigated by use of structural equation models in a cross-lagged panel design. The theory of reasoned action and an augmented version of the theory, which included the frequency and recency of past behavior as covariates, were tested. Respondents were 254 undergraduates who provided behavioral and psychological reactions over two points in time to two kinds of goal-directed behaviors: event-planned goals (i.e., trying to lose weight) and event-triggered goals (i.e., initiating a conversation with an attractive stranger). The findings show that the theory of reasoned action as classically formulated explains trying to lose weight well but initiating a conversation poorly. Further, the introduction of frequency and recency effects brings into question predictions under the theory of reasoned action. Theoretical implications of the results are discussed. A weighted least squares method is applied to the asymptotic covariance matrix based on appropriate polychoric and tetrachoric correlations. Variates are assumed measured on only ordinal scales.
在交叉滞后面板设计中，通过使用结构方程模型来研究态度与行为关系的本质。对理性行动理论以及该理论的一个扩充版本进行了检验，扩充版本将过去行为的频率和近期性作为协变量纳入其中。研究对象为254名本科生，他们在两个时间点针对两种目标导向行为提供了行为和心理反应：事件计划目标（即试图减肥）和事件触发目标（即与有吸引力的陌生人展开对话）。研究结果表明，经典表述的理性行动理论能很好地解释试图减肥的行为，但对展开对话行为的解释效果不佳。此外，频率和近期效应的引入对理性行动理论下的预测提出了质疑。文中讨论了研究结果的理论意义。基于适当的多系列相关和四分相关，将加权最小二乘法应用于渐近协方差矩阵。假设变量仅在有序尺度上进行测量。
