Researchers often study constructs that are conceptually and/or empirically related, but distinct (i.e., “sibling constructs”). In social-personality psychology, as well as psychology more generally, there is little guidance for how to deal with sibling constructs, which can result in researchers ignoring or mishandling them. In this article, we start by situating sibling constructs in the literature on the jingle-jangle fallacies. Then, we outline 10 conceptual and empirical criteria for determining the degree to which, and in what ways, constructs may share a sibling relationship, using self-esteem and grandiose narcissism as a running example. Finally, we discuss strategies for handling sibling constructs in a systematic and transparent way. We hope that the procedures described here will help social-personality psychologists identify sibling constructs, understand when and why they pose problems for their research, and adopt strategies that ameliorate their adverse effects.

It is common practice in correlational or quasiexperimental studies to use statistical control to remove confounding effects from a regression coefficient. Controlling for relevant confounders can debias the estimated causal effect of a predictor on an outcome; that is, it can bring the estimated regression coefficient closer to the value of the true causal effect. But statistical control works only under ideal circumstances. When the selected control variables are inappropriate, controlling can result in estimates that are more biased than uncontrolled estimates. Despite the ubiquity of statistical control in published regression analyses and the consequences of controlling for inappropriate third variables, the selection of control variables is rarely explicitly justified in print. We argue that to carefully select appropriate control variables, researchers must propose and defend a causal structure that includes the outcome, predictors, and plausible confounders. We underscore the importance of causality when selecting control variables by demonstrating how regression coefficients are affected by controlling for appropriate and inappropriate variables. Finally, we provide practical recommendations for applied researchers who wish to use statistical control.

Victims to Victory We investigated the psychological impact of homicide bereavement in a sample of 54 African Americans (88.9% female) who had experienced the murder of a loved one within the past 5 years. Participants were recruited from a victims' services agency. The majority of participants (n ϭ 34, 63%) were parents of the deceased. Using a cutoff of 50 on the PTSD Checklist (Weathers, Litz, Herman, Huska, & Keane, 1993), 10 participants (18.5%) screened positive for PTSD. On the Beck Depression Inventory-II, 54% of the sample (n ϭ 29) had scores suggesting at least mild depression. On the Inventory of Complicated Grief (Prigerson & Jacobs, 2001), 24 (54.5%) of those for whom the homicide occurred 6 months or more prior to assessment screened positive for complicated grief. There was a high degree of overlap across these categories, with nearly all of the PTSD-positive cases screening positive for complicated grief and depression. Participants who were within 2 years of a homicide loss showed significantly higher levels of PTSD and anxiety severities than those who were 2 or more years beyond the loss. In contrast, levels of complicated grief and depression did not differ significantly between those early and late in bereavement. In regression analyses, time since the homicide was a significant predictor for anxiety and approached significance in predicting PTSD. However, time since homicide was not significantly associated with depression or complicated grief. Clinical and research implications of these findings are discussed, including the possible impact of stigma associated with homicidal bereavement.

Experience programs synaptic development to match the needs of the environment. This process depends on the nature and timing of the experience. Exposure to stress during adolescence selectively reduces synaptic density in the prefrontal cortex (a later maturing region), while sparing hippocampal synapses (an earlier maturing region). To determine whether the anatomical effects of an adolescent social stressor in rats endures into adulthood and are reversible, male subjects were isolation or group housed between days 30 and 35 and then treated with vehicle, adinazolam, MK-801, or tianeptine between days 40 and 55. At day 60, immunohistochemistry revealed a 13.5% +/- 5.3% reduction in synaptophysin in the infralimbic cortex and cingulate gyrus in isolation-housed subjects. MK-801 and adinazolam restored cortical synaptic density to within 2% of group-housed values, suggesting that the synaptic loss induced by stress during adolescence is modulated through reduced glutamatergic activity directly by NMDA antagonism or indirectly by enhancing GABAergic activity. Tianeptine did not modulate adolescent stress effects in the prefrontal cortex. None of these drugs increased cortical synaptophysin in group-housed controls. Increased synaptophysin was observed in the group-housed condition in the hippocampus, striatum, and nucleus accumbens following drug exposure. Although stress did not decrease synaptic density in these regions, drug exposure failed to increase synaptic density when compared with the controls. Taken together, stress-induced changes in cortical, but not hippocampal, synaptic density initiated during adolescence endure into adulthood. These cortical changes can be reversed through a reduction of glutamatergic activity, but not serotonin augmentation.