Semiparametric Efficient Estimation in the Generalized Odds-Rate Class of Regression Models for Right-Censored Time-to-Event Data

The generalized odds-rate class of regression models for time to event data is indexed by a non-negative constant and assumes thatg(S(t|Z)) = (t) + Zwhere g(s) = log(^-1(s-) for > 0, g₀(s) = log(- log s), S(t|Z) is the survival function of the time to event for an individual with qx1 covariate vector Z, is a qx1 vector of unknown regression parameters, and (t) is some arbitrary increasing function of t. When =0, this model is equivalent to the proportional hazards model and when =1, this model reduces to the proportional odds model. In the presence of right censoring, we construct estimators for and exp((t)) and show that they are consistent and asymptotically normal. In addition, we show that the estimator for is semiparametric efficient in the sense that it attains the semiparametric variance bound.     

On Tests for Group Variation With a Small to Moderate Number of Groups

This paper considers a family of penalized likelihood score tests for group variation. The tests can be indexed by a measure of degrees of freedom. At one extreme, with degrees of freedom one less than the number of groups, is the usual score test for a fixed effects alternative using indicator variables for the groups, while at the other extreme, in the limit as the degrees of freedom 0, is a test closely related to a score test based on a random effects alternative. Asymptotic power comparisons are made for the tests in the family. As would be expected, different members of the family are more efficient for different alternatives. Generally the tests with smaller degrees of freedom appear to have better power than the standard test for alternatives focusing on differences among the larger groups, and lower power for alternatives focusing on differences among the smaller groups. Simulations indicate the asymptotic approximation to the distribution performs better for the tests with small degrees of freedom.     

A panel study of migration,self-selection and household real income

The impact of migration on income for Swedish multi-adult households is examined using panel data pertaining to a sample of stable household constellations during the period 1980–1990. In contrast to previous studies, data on household disposable income is employed in estimating the income function. The empirical results indicate no significant effect on real disposable income from migration. In addition, the hypothesis of no self-selection, or zero correlation between the errors in the decision function and the income function, cannot be rejected. Received: 10 May 1995 / Accepted: 14 August 1996     

An Analysis of Suicide Rates among Occupational Categories

The correlation between the NORC prestige scale of 1947 with 1950 suicide rates is extremely low for thirty-six specific occupations. The low correlation calls into question the findings of previous studies which employed gross occupational categories and impressionistic prestige rankings. The results indicate that client dependency and two multivariate models, status inconsistency and distributive justice, deserve further exploration in predicting suicide rates for occupations.     

An analysis of vendor innovation capability in the contract electronics manufacturing industry

Limited academic research has been given to analysing the innovation capabilities of vendors in outsourcing contracts. This paper seeks to address this gap in the literature by enhancing our understanding of how the innovation capability of vendors is deployed to win, run and renew outsourcing contracts with their customers. Employing the resource-based view as a theoretical basis and undertaking in-depth case study analysis of three vendors in the electronic manufacturing services industry, the research shows that to achieve the outsourcing objectives of winning, running and renewing the contract, vendors can use different configurations of the competitive priorities of cost, quality, delivery and flexibility. The research aggregates the capabilities that influence the innovative capability of a vendor into the innovation-related capabilities (IRCs) of design, new product introduction and manufacturing. Three strategies are identified for vendors on how to deploy these IRCs, and a number of propositions are developed to indicate the suitability of the three deployment strategies for different operational contexts.     

Response to Bonanno and Nehring

Theory and Decision -     

Disaggregating the Motives Underlying Turnover Intentions: When Do Intentions Predict Turnover Behavior?

Although intention to quit an organizationrepresents one of the better predictors of actualturnover behavior, the intention behavior relationshipvaries widely across studies. This study argues that one reason for the variability is that individualspossess different motives for stating a high intentionof leaving, and it is the motive that accounts in partfor when the intention will manifest itself in actual turnover. Using a longitudinal design,expectations were supported in that only thoseintentions motivated by disaffection with theorganization and its values resulted in the loss ofemployees.     

Urbanization, Disadvantage, and Petty Entrepreneurship: Street Peddling among African American Men in Southern and Northern Cities during the Early Twentieth Century

Two theories attempt to explain why minority groups are often involved in small-scale entrepreneurial activities. The urban adjustment hypothesis holds that when such groups urbanize, their members establish themselves economically through self-employment in pursuits that require little start-up capital. The disadvantage theory argues that exclusion from the labor market impels members of oppressed groups to become self-employed in marginal enterprises. Applying these theories, I use census data to analyze the participation of African American men in street peddling during the pre-World War II era. Consistent with the urban adjustment hypothesis, the involvement of African American men in street peddling increased markedly when African Americans urbanized during the period from 1890 to 1940. Consistent with disadvantage theory, the involvement of African American men in street peddling was positively associated with labor market disadvantage. Finally, in southern cities, African American men were overrepresented in street peddling, a finding that suggests that these men had a niche in the petty-trading sector of the South's informal economy.     

MARITAL THERAPY: QUALITIES OF COUPLES WHO FARE BETTER OR WORSE IN TREATMENT

To evaluate the effectiveness of couples therapy and determine the characteristics and qualities of couples who fare best in treatment, 139 couples presenting at a nonlaboratory sliding-fee clinic in Dallas, Texas, were evaluated. Observational and self-report measures of the Beavers Systems Model were taken at the first session and, for some subjects, at termination. Therapists also rated the couples on goal attainment and on dimensions of the therapeutic alliance. Not surprisingly, there was a direct association between number of sessions and goal attainment, with couples attending only one or two sessions having the lowest goal attainment. There were also significant family functioning qualities associated with success: more competent couples fared better in therapy. Another consistent finding was that couples with no children did better in treatment than those who were also parents. It is interesting that remarried couples without children did best in treatment, followed by first-married couples with no children, first-married couples with children, and, finally, remarried couples with children. Implications for screening and clinical work with couples are presented.