Causal graphical views of fixed effects and random effects models

Despite the long-standing discussion on fixed effects (FE) and random effects (RE) models, how and under what conditions both methods can eliminate unmeasured confounding bias has not yet been widely understood in practice. Using a simple pretest–posttest design in a linear setting, this paper translates the conventional algebraic formalization of FE and RE models into causal graphs and provides intuitively accessible graphical explanations about their data-generating and bias-removing processes. The proposed causal graphs highlight that FE and RE models consider different data-generating models. RE models presume a data-generating model that is identical to a randomized controlled trial, while FE models allow for unobserved time-invariant treatment–outcome confounding. Augmenting regular causal graphs that describe data-generating processes by adding the computational structures of FE and RE estimators, the paper visualizes how FE estimators (gain score and deviation score estimators) and RE estimators (quasi-deviation score estimators) offset unmeasured confounding bias. In contrast to standard regression or matching estimators that reduce confounding bias by blocking non-causal paths via conditioning, FE and RE estimators offset confounding bias by deliberately creating new non-causal paths and associations of opposite sign. Though FE and RE estimators are similar in their bias-offsetting mechanisms, the augmented graphs reveal their subtle differences that can result in different biases in observational studies.     

Exploiting the Wisdom of Others to Make Better Decisions: Suspending Judgment Reduces Egocentrism and Increases Accuracy

Although decision makers often consult other people's opinions to improve their decisions, they fail to do so optimally. One main obstacle to incorporating others' opinions efficiently is one's own opinion. We theorize that decision makers could improve their performance by suspending their own judgment. In three studies, participants used others' opinions to estimate uncertain quantities (the caloric value of foods). In the full‐view condition, participants could form independent estimates prior to receiving others' opinions, whereas participants in the blindfold condition could not form prior opinions. We obtained an intriguing blindfold effect. In all studies, the blindfolded participants provided more accurate estimates than did the full‐view participants. Several policy‐capturing measures indicated that the advantage of the blindfolded participants was due to their unbiased weighting of others' opinions. The full‐view participants, in contrast, adhered to their prior opinion and thus failed to exploit the information contained in others' opinions. Moreover, in all three studies, the blindfolded participants were not cognizant of their advantage and expressed less confidence in their estimates than did the full‐view participants. The results are discussed in relation to theories of opinion revision and group decision making. Copyright © 2011 John Wiley & Sons, Ltd.     

Estimating the causal effect of randomization versus treatment preference in a doubly randomized preference trial

Although randomized studies have high internal validity, generalizability of the estimated causal effect from randomized clinical trials to real-world clinical or educational practice may be limited. We consider the implication of randomized assignment to treatment, as compared with choice of preferred treatment as it occurs in real-world conditions. Compliance, engagement, or motivation may be better with a preferred treatment, and this can complicate the generalizability of results from randomized trials. The doubly randomized preference trial (DRPT) is a hybrid randomized and nonrandomized design that allows for estimation of the causal effect of randomization versus treatment preference. In the DRPT, individuals are first randomized to either randomized assignment or choice assignment. Those in the randomized assignment group are then randomized to treatment or control, and those in the choice group receive their preference of treatment versus control. Using the potential outcomes framework, we apply the algebra of conditional independence to show how the DRPT can be used to derive an unbiased estimate of the causal effect of randomization versus preference for each of the treatment and comparison conditions. Also, we show how these results can be implemented using full matching on the propensity score. The methodology is illustrated with a DRPT of introductory psychology students who were randomized to randomized assignment or preference of mathematics versus vocabulary training. We found a small to moderate benefit of preference versus randomization with respect to the mathematics outcome for those who received mathematics training.     

The delocalized mind. Judgements,vehicles, and persons

Drawing on various resources and requirements (as expressed by Dewey, Wittgenstein, Sellars, and Brandom), this paper proposes an externalist view of conceptual mental episodes that does not equate them, even partially, with vehicles of any sort, whether the vehicles be located in the environment or in the head. The social and pragmatic nature of the use of concepts and conceptual content makes it unnecessary and indeed impossible to locate the entities that realize conceptual mental episodes in non-personal or subpersonal contentful entities (vehicles). Persons, who engage in social practices of deontic scorekeeping (Brandom), and who are thus able to produce appropriate inferential behaviour, enact conceptual mental episodes and the basis of their supervenience.     

Seeing and being seen: Shame in the clinical situation

Shame may prevent the patient from emerging from a psychic retreat. As begins to do so he confronts two fears, first of seeing the object more clearly and second of being seen become prominent. Seeing leads to deeper and more distressing feelings connected with guilt and depression as the damage done to good objects is recognized. However it cannot be faced if shame leads to a demand for immediate relief. Shame is a prominent feature of the analytic situation and recognizing this may help the analyst to support his patients to tolerate the discomfort of being seen so that the conflicts about seeing can be worked through. Two clinical examples are briefly discussed. In the first feelings of inferiority lessened as they were analysed and allowed appreciative and depressive feelings to emerge. In the second embarrassment was associated with progress that the patient felt he had made but was embarrassed to admit. It is argued that the analysis of shame in the analytic situation is necessary so that being seen can be tolerated and allow the conflicts over seeing to be worked through.     

Effects of motherless rearing on basal and stress-induced corticosterone secretion in rat pups

Rearing of rat pups without a mother, artificial rearing (AR), produces substantial changes in the pups' behavior in later life. These changes are similar to those produced by the stress of repeated mother-pup separations. The predominant interpretation is that the long-term effects of disruptions to the mother-pup relationship are mediated by exposure to elevated levels of corticosterone which affect the development of neurobiological systems underlying cognition and behavior. Indeed, repeated separation of pups from the mother sensitizes the pups' corticosterone response to stress. This study examined basal and stress-induced corticosterone release in AR pups. Corticosterone levels were increased immediately following implantation of feeding cannulae. One day after the start of AR, circulating concentrations of corticosterone were not increased unless AR pups were challenged with an additional stressor (injection). Corticosterone levels were lowest when cannulation and AR started on postnatal day (PND) 5 compared with earlier PNDs. On PND 12, there was no evidence of increased corticosterone levels in AR pups at baseline or in response to stress, indicating that AR did not result in persistent sensitization of corticosterone release. The long-term effects of motherless rearing on rat behavior are mediated by mechanisms that are independent of sustained early corticosterone exposure.