Assessment of Multiple Membership Multilevel Models: An Application to Interviewer Effects on Nonresponse

Multilevel multiple membership models account for situations where lower level units are nested within multiple higher level units from the same classification. Not accounting correctly for such multiple membership structures leads to biased results. The use of a multiple membership model requires selection of weights reflecting the hypothesized contribution of each level two unit and their relationship to the level one outcome. The Deviance Information Criterion (DIC) has been proposed to identify such weights. For the case of logistic regression, this study assesses, through simulation, the model identification rates of the DIC to detect the correct multiple membership weights, and the properties of model variance estimators for different weight specifications across a range of scenarios. The study is motivated by analyzing interviewer effects across waves in a longitudinal study. Interviewers can substantially influence the behavior of sample survey respondents, including their decision to participate in the survey. In the case of a longitudinal survey several interviewers may contact sample members to participate across different waves. Multilevel multiple membership models are suitable to account for the inclusion of higher-level random effects for interviewers at various waves, and to assess, for example, the relative importance of previous and current wave interviewers on current wave nonresponse. To illustrate the application, multiple membership models are applied to the UK Family and Children Survey to identify interviewer effects in a longitudinal study. The paper takes a critical view on the substantive interpretation of the model weights and provides practical guidance to statistical modelers. The main recommendation is that it is best to specify the weights in a multiple membership model by exploring different weight specifications based on the DIC, rather than prespecifying the weights.     

Born to kill? A critical evaluation of homicide adaptation theory

Given the enormous harm that results from the intentional killing of humans by other humans, the development of etiological theories of homicide is an important task. Homicide Adaptation Theory (HAT) is one such attempt to explain why homicide occurs. In short, the proponents of HAT propose that humans possess a number of specific, evolved adaptations for killing that have been selected for because they managed to successfully solve recurrent adaptive problems in our ancestral past. In this paper it is argued that, although evolutionary approaches can potentially advance our understanding of homicide in important ways, the available evidence does not support the existence of specific adaptations for killing in humans. Specifically, I argue that the proponents of HAT have not provided: (a) a thorough analysis of the evolutionary costs and benefits of killing; (b) evidence for ‘special design’ features of homicide; or (c) a systematic comparative appraisal of killing among different species. Given this lack of evidence, HAT suffers in terms of its explanatory worth, relative to alternative evolutionary hypotheses, such as the idea that homicide is a by-product of other evolved adaptations.     

Contributions of impulsivity and callousness in the assessment of adjudicated male adolescents: a prospective study

This study examined whether delinquent adolescents with low scores on personality scales related to callousness and impulsivity would show less antisocial behavior and better psychosocial functioning compared to those elevated on 1 or both scales. A group of 162 White and African American male adolescents, recently adjudicated in a southern youth court system, were assessed at baseline and at 18-month follow-up. Dependent variables included measures of antisocial and prosocial behaviors, days detained in juvenile detention, and symptoms of psychopathology. As hypothesized, adolescents low on both callousness and impulsivity spent fewer days detained and reported less antisocial behavior and fewer symptoms of psychopathology compared to those high on I or both scales. Finally, African American participants were detained for more days, but did not report significantly more antisocial behaviors or symptoms of psychopathology compared to the White adolescents. The results suggest that being low on callousness and impulsivity may serve a protective role for future delinquent behavior for both African American and White adolescents.     

Discomfort with space and motion: A possible marker of vestibular dysfunction assessed by the situational characteristics questionnaire

Space and motion discomfort (SMD) refers to the situational specificity of symptoms occurring in some patients with vestibular dysfunction, such as those with balance disorders and some with panic disorder. SMD occurs in situations characterized by inadequate visual or kinesthetic information for normal spatial orientation. We report the results of two studies of the construct validity of the Situational Characteristics Questionnaire (SitQ), which has two subscales, both of which measure SMD: the SMD-I and SMD-II. In Study 1, the SitQ was administered to members of a self-help group for balance disorders, a psychiatric sample consisting of patients with panic disorder, nonpanic anxiety disorders, depression, and a sample of normals. SMD levels were the highest in the self-help balance group, next to the highest in the panic groups, and lowest in the remaining groups. In Study 2, the SitQ was administered to otolaryngological patients with vestibular dysfunction and to patients with hearing loss. SMD levels were higher in the vestibular patients. Data on internal consistency, test-retest reliability, and convergent and discriminant validity are presented. The SitQ, particularly the SMD-II, is recommended for quantifying space and motion discomfort in patients with anxiety and/or balance disorders.Study 1 was funded by MH 40757. Study 2 was funded by a grant from the Upjohn Company and MH 19816.