Extending mixtures of multivariate <Emphasis Type="Italic">t</Emphasis>-factor analyzers

Model-based clustering typically involves the development of a family of mixture models and the imposition of these models upon data. The best member of the family is then chosen using some criterion and the associated parameter estimates lead to predicted group memberships, or clusterings. This paper describes the extension of the mixtures of multivariate t-factor analyzers model to include constraints on the degrees of freedom, the factor loadings, and the error variance matrices. The result is a family of six mixture models, including parsimonious models. Parameter estimates for this family of models are derived using an alternating expectation-conditional maximization algorithm and convergence is determined based on Aitken’s acceleration. Model selection is carried out using the Bayesian information criterion (BIC) and the integrated completed likelihood (ICL). This novel family of mixture models is then applied to simulated and real data where clustering performance meets or exceeds that of established model-based clustering methods. The simulation studies include a comparison of the BIC and the ICL as model selection techniques for this novel family of models. Application to simulated data with larger dimensionality is also explored.     

The validity of measures of self-reported well-being

Using a new analytic approach, construct validity estimates are developed for proposed social indicators of self-reported well-being. Two separate investigations are reported: the first involves data on six aspects of well-being each assessed by six methods from 222 adults in one geographic area; the second, a partial replication and extension, involves a more limited set of indicators measured on a sample of 1297 respondents representative of all American adults. The results provide evidence that perceptions of well-being can be measured by single questionnaire or interview items using any of four formats with validities in the range of 0.7 to 0.8 and with correlated method effects contributing less than 10% of the total variance. Two other formats, however, were markedly less valid. These findings are important in view of past criticisms of ‘subjective’ social indicators as lacking in validity, and the findings can guide current efforts to develop new ways to assess the quality of life. Methodologically, the article illustrates the feasibility and utility of deriving parameter estimates of structural equation models of multimethod-multitrait data using Joreskog's LISREL algorithm. The possibility of deriving validity estimates in this way, even when the data include correlated errors, opens new and important opportunities to precisely assess the amount of error variance in much social science data.     

成功老龄化研究--以北京老年人为例

本文利用北京老龄化多维纵向研究数据,对北京老年人的成功老龄化状况进行了分析.分析结果表明,年龄、性别、受教育程度和家庭经济状况等因素影响着成功老龄化.属于成功老龄化范畴的老年人主要是低年龄、男性、受教育时间长、拥有较多家庭财产或月收入的老年人.上述因素并不只对某一个年龄组产生影响,它们对老年的各个阶段都有影响.年龄的增长与老年人身体、心理和社会功能有着紧密的联系,在控制了年龄、性别、教育和收入变量之后,老年人之间的差异依然存在.     

Explaining Career Decision‐Making Self‐Efficacy: Personality,Cognitions, and Cultural Mistrust

The authors explore the hypothesis that career decision‐making self‐efficacy could be affected by negative career thoughts, Big Five personality factors, and cultural mistrust in a sample of African American and Caucasian college students. Findings demonstrated that negative career thinking, openness, and conscientiousness explained a significant amount of variance in career decision‐making self‐efficacy in a general sample of college students, but no unique variance was explained by cultural mistrust in a sample of African American college students.     

Organizational Form and Strategic Alignment in a Local Authority: A Preliminary Exploration using Fuzzy Clustering

Configurational theories assume that organizational form has important implications for the degree of alignment between top and middle management on strategic priorities. Taken in combination, the structure, process and environment of an organization are thought to have a deep pervasive influence on top management’s attempts to achieve the coordination and control required to attain organizational goals. The preliminary analysis described in this article employs fuzzy c-means clustering to explore the relationship between middle managers’ perceptions of organizational form and strategic alignment within a large local authority. The results illustrate that the clustering of managers’ perceptions of organizational structure, process and environment reflect three organizational archetypes: machine bureaucracy, professional bureaucracy and professional adhocracy. Statistically significant differences in the degree of strategic alignment between each of these organizational forms are then examined to validate the established clustering. Finally, conclusions are drawn on the theoretical and practical implications of the findings.     

Stress from infertility, marriage factors, and subjective well-being of wives and husbands

Stress associated with the inability to have a child is linked to four aspects of marriage and to five dimensions of life quality. Data come from 157 couples who met a standard medical definition for infertility. Wives and husbands were interviewed independently, and most of the following findings apply to both. A causal model suggests that fertility problem stress has direct effects that increase marital conflict and decrease sexual self-esteem, satisfaction with own sexual performance, and frequency of sexual intercourse. Also, fertility problem stress has both direct and indirect effects (via the marriage factors) that decrease evaluations of life-as-a-whole, self-efficacy, marriage, intimacy, and health. The negative effects on life quality are stronger for wives than for husbands. The model suggests that the life quality of couples with fertility problems could be improved if health care providers and couples themselves took steps to reduce such stresses and/or reduce their impact on the marriage factors.     

Assessing Perceptions of Synergistic Health Risk: A Comparison of Two Scales

Two approaches to measuring perceptions of synergistic risk were compared, one using the traditional Likert scale, the other using an anchored, relative scale. Perception of synergistic risk was defined as rating the combined hazard as more risky than each of its constituent single hazards. In a within-subjects design, a convenience sample from the community (N= 604) rated three hazard combinations and their constituents: Driving while Intoxicated (familiar, high synergy), Radon and Smoking (unfamiliar, high synergy), and Smoking and Driving (familiar, low synergy), on both scales. The relative scale was expected to be a more sensitive measure of synergy than the Likert scale. The effects of item order (single hazards rated first versus combined hazards rated first) were examined between subjects. Driving while Intoxicated was perceived by the large majority of participants as a synergistic risk on both scales, but neither of the other two combined hazards were rated synergistically on either scale. The relative scale demonstrated a slight advantage over the Likert scale, and presenting the single hazards first for the relative scale produced more synergistic ratings. It is recommended that anchored, relative scales be used to measure synergy and that single hazards be presented prior to the combined hazards when using relative scales.     

Subjective Reactions to Two Regional Pronunciations of Great Russian: A Matched-Guise Study

En sociolinguistique russe, il est un domaine bien négligé, celui de l’examen impartial des préjugés linguistiques contre les variétés de langue non-standard. Dans cette étude, j’ai travaillé avec des sujets russes parlant un russe standard et ayant reçu une éducation supérieure, pour mesurer leurs réactions subjectives aux deux prononciations régionales les plus marquantes du grand-russe: Vokan’e où la voyelle [o] est maintenue dans des syllabes non-accentuées, et la prononciation russe du sud dans laquelle l’occlusive vélaire voisée [g] est remplacée par la consonne fricative [Y] ou aspirée [h]. La méthode expériméntale utilisée a été la technique “Matched-Guise” dans laquelle les sujets écoutent une série de voix enregistrées et les évaluent chacune pour divers attributs sur une échelle numérique. Certains des sujets enregistrés l’ont été dans les deux varieétés de langue, la langue de prestige et la langue stigmatisée, ce qui permet de voir les réactions aux différences dans la qualité de la voix. Les données recueillies montrent que les prononciations non-standard du grand-russe sont dévalorisées pour des attributs liés au statut, et à cet égard, le russe est semblable à d’autres langues européennes qui ont une langue standard extrêmement codifiée. On a également constaté, ce qui est plus inhabituel, l’importante revalorisation des sujets parlant Vokan’e pour des attributs personnels, ainsi qu’une dévalorisation marquée de la prononciation du sud pour ées mêmes catégories. Cette dichotomie doit résulter de conditions spécifiques à l’ex-Union Soviétique. La réaction positive à l’ okan’e, un dialecte rural selon le stéréotype, peut avoir été inspirée par une littérature de village en prose, alors que la réaction négative à la prononciation du sud est peut-être liée aux dirigeants politiques de l’ère soviétique qui l’ont utilisée. Ces données confirment également l’efficacité de la technique “Matched-Guise” pour conduire ce type ?’enquête sociolinguistique en Russie.