Sensitivity Analysis in Structural Equation Models: Cases and Their Influence

The detection of outliers and influential observations is routine practice in linear regression. Despite ongoing extensions and development of case diagnostics in structural equation models (SEM), their application has received limited attention and understanding in practice. The use of case diagnostics informs analysts of the uncertainty of model estimates under different subsets of the data and highlights unusual and important characteristics of certain cases. We present several measures of case influence applicable in SEM and illustrate their implementation, presentation, and interpretation with two empirical examples: (a) a common factor model on verbal and visual ability (Holzinger &; Swineford, 1939 Holzinger, K. and Swineford, F. 1939. A study in factor analysis: The stability of a bi-factor solution. Chicago, IL: University of Chicago..  [Google Scholar]) and (b) a general structural equation model assessing the effect of industrialization on democracy in a mediating model using country-level data (Bollen, 1989 Bollen, K. A. 1989. Structural equation models with latent variables New York, NY: Wiley.. [Crossref] , [Google Scholar]; Bollen &; Arminger, 1991 Bollen, K. A. and Arminger, G. 1991. Observational residuals in factor analysis and structural equation models. Sociological Methodology, 21: 235–262. [Crossref], [Web of Science ®] , [Google Scholar]). Throughout these examples, three issues are emphasized. First, cases may impact different aspects of results as identified by different measures of influence. Second, the important distinction between outliers and influential cases is highlighted. Third, the concept of good and bad cases is introduced—these are influential cases that improve/worsen overall model fit based on their presence in the sample. We conclude with a discussion on the utility of detecting influential cases in SEM and present recommendations for the use of measures of case influence.     

Humean dispositionalism

Humean metaphysics is characterized by a rejection of necessary connections between distinct existences. Dispositionalists claim that there are basic causal powers. The existence of such properties is widely held to be incompatible with the Humean rejection of necessary connections. In this paper I present a novel theory of causal powers that vindicates the dispositionalist claim that causal powers are basic, without embracing brute necessary connections. The key assumptions of the theory are that there are natural types of causal processes, and that manifestations of powers are identified with certain kinds of causal processes. From these assumptions, the modal features of powers are explained in terms of internal relations between powers themselves and the process-types in which powers are manifested.     

Discrimination, dispositions, and cardiovascular responses to stress.

OBJECTIVE: Recent research suggests that past exposure to discrimination may influence perceptions of, and physiological responses to, new challenges. The authors examined how race and trait levels of hostility and optimism interact with past exposure to discrimination to predict physiological reactivity and recovery during an anger recall task. DESIGN: A community sample of 165 normotensive Black and White adults participated in an anger recall task while having their cardiovascular function monitored. MAIN OUTCOME MEASURES: Blood pressure and heart rate indicators of physiological reactivity and recovery. RESULTS AND CONCLUSION: Participants had higher reactivity and slower recovery to the anger recall task when they had high past discrimination, low cynicism, or high optimism. The pattern of effects was similar for both racial groups, but Blacks had more acute reactivity and slower recovery than Whites. These results are consistent with the perspective of discrimination as a chronic stressor that is related to acute stress responses, particularly for Blacks.     

Relevant Restricted Quantification

The paper reviews a number of approaches for handling restricted quantification in relevant logic, and proposes a novel one. This proceeds by introducing a novel kind of enthymematic conditional.     

Domain-specific assessment of anger expression and ambulatory blood pressure

Bongard and al’Absi (2003) proposed domain-specific anger expression as an improved method for measuring the expression of anger with stronger links to cardiovascular parameters. We tested this proposal by relating their domain-specific measure to ambulatory blood pressure. One hundred and forty-nine Singapore young adults responded to a modified version of the State-Trait Anger Expression Inventory that measured anger expression at home, school/work and leisure and then underwent 24 h ambulatory blood pressure monitoring. Results indicated significant differences in reported anger expression in the three domains measured and also showed that domain-specific measures of anger expression were more strongly related to ambulatory blood pressure than was the general measure of anger expression. These results provide additional evidence for the importance of measures of anger expression that take account of the specific context in which anger occurs.