BAYESIAN SUBSET SELECTION AND MODEL AVERAGING USING A CENTRED AND DISPERSED PRIOR FOR THE ERROR VARIANCE

This article proposes a new data‐based prior distribution for the error variance in a Gaussian linear regression model, when the model is used for Bayesian variable selection and model averaging. For a given subset of variables in the model, this prior has a mode that is an unbiased estimator of the error variance but is suitably dispersed to make it uninformative relative to the marginal likelihood. The advantage of this empirical Bayes prior for the error variance is that it is centred and dispersed sensibly and avoids the arbitrary specification of hyperparameters. The performance of the new prior is compared to that of a prior proposed previously in the literature using several simulated examples and two loss functions. For each example our paper also reports results for the model that orthogonalizes the predictor variables before performing subset selection. A real example is also investigated. The empirical results suggest that for both the simulated and real data, the performance of the estimators based on the prior proposed in our article compares favourably with that of a prior used previously in the literature.     

Higher education outcomes, graduate employment and university performance indicators

Summary.  Official employment-related performance indicators in UK higher education are based on the population of students responding to the 'First destination supplement' (FDS). This generates potentially biased performance indicators as this population of students is not necessarily representative of the full population of leavers from each institution. University leavers who do not obtain qualifications and those who do not respond to the FDS are not included within the official analysis. We compare an employment-related performance indicator based on those students who responded to the FDS with alternative approaches which address the potential non-random nature of this subgroup of university leavers.     

Fundamentals of Theory and Practice Revisited; The Mutiny and the Bounty: The Place of Milan Ideas Today

The author traces the development of the Milan approach from the beginning, through the split into two teams in 1980, and subsequent developments in theory and application. He identifies possible unrecognised or undeclared influences of Milan in other family therapy approaches.     

The effects of language on judgments of universal facial expressions of emotion

Because of the close connection between culture and language, a number of writers have suggested that bilinguals will differ in their behavior because of differences in the degree of assimilation of different cultures in the same individual. We tested this notion by obtaining data from bilingual (English and Hindi) college students in India using a well-studied cross-cultural research paradigm involving emotional perception. Subjects judged universal facial expressions of emotion in two separate sessions, one conducted entirely in English, the other in Hindi. In each session, they judged which emotion was being portrayed, and how intensely. Subjects recognized anger, fear, and sadness more accurately in English than in Hindi. They also attributed greater intensity to female photos of anger when rating in Hindi, but attributed greater intensity to female photos of sadness when rating in English. These findings were discussed in relation to the theoretical connection between culture and language.     

Vulnerability and Agenda: Context and Process in Project Management

This paper explores how the project management role is conditioned by the context in which change is progressed. The argument draws on Pettigrew's (1985, 1987) contextual and processual view of change, a view which does not clarify the management implications of contextual variation. Using data from a two-stage research design combining diary and survey methods, four context dimensions are identified, concerning ‘interlocking’, ‘shifting sands’, ‘ownership’ and ‘senior management view’. For analytical purposes, two extreme contexts are characterized as exposing the project manager to high and low levels of ‘vulnerability’ respectively. Project management literature typically offers a rational-linear account of change, and concentrates on ‘content’ and ‘control’ agendas, concerning technical expertise on the one hand, and planning, budgeting and monitoring techniques on the other. Sociological analyses reveal the limitations of the rational-linear account, and focus on the political and cultural dimensions of the ‘process’ agenda. The data suggest how contextual variation affects the relative priority of these agendas. The management implications of these findings are explored, for project management selection, career progression, and the development of diagnostic skills and ‘agenda management’ strategies.     

Ethnic preferences and perceptions among Asian and White British middle school children

Asian and White middle school children were shown photographs of unknown Afro-Caribbean, Asian, and White individuals of approximately the same age as themselves, and asked which one wax most like them, which one they would prefer to be, which ones they would prefer to engage in various activities with, and finally, to allocate positive and negative traits to the three groups or to nobody. A similar majority of both Asian and While children selected an own-race photograph as being most like them, but only about a half selected the own-race photograph as the one they would most like to he. For sharing activities, children showed a strong own-tender preference, followed by an own-race preference. White children stated a greater preference for sharing activities with White children, then with Afro-Caribbean children, and least with Asian children. Asian children preferred Asian and White children the most, Afro-Caribbean the least. On a stereotypes test, similar trends were obtained; White children evaluated While children highest and Asian lowest; Asian children tended to evaluate Asian children highest but to a nonsignificant extent. The relative proportion of Asian/White children in school appeared to have little systematic influence on the results, thus arguing against the contact hypothesis of prejudice reduction in its simple form.     

PREDICTING MARITAL SUCCESS FOR PREMARITAL COUPLE TYPES BASED ON PREPARE

Recent studies have shown that group differences in marital stability and satisfaction can be predicted based on premarital relationship quality. There is also a growing literature indicating that there are distinct types of relationships, both premaritally and developmentally over time. This study examined the relationship between the four premarital types (Vitalized, Harmonious, Traditional, and Conflicted) identified by Fowers and Olson (1992) and relationship outcome over a 3-year period with 393 couples. A substantial relationship was found, with conflicted couples being the most likely to separate or divorce. Vitalized couples had the highest levels of satisfaction, followed by Harmonious, Traditional, and Conflicted couples. Traditional couples were less likely to have divorced than Harmonious couples, even though Harmonious couples had higher premarital relationship satisfaction scores.