Out with .05, in with Replication and Measurement: Isolating and Working with the Particular Effect Sizes that are Troublesome for Inferential Statistics

It is difficult to obtain adequate power to test a small effect size with a set criterion alpha of 0.05. Probably an inferential test will indicate non-statistical significance and not be published. Rarely, statistical significance will be obtained, and an exaggerated effect size calculated and reported. Accepting all inferential probabilities and associated effect sizes could solve exaggeration problems. Graphs, generated through Monte Carlo methods, are presented to illustrate this. The first graph presents effect sizes (Cohen's d) as lines from 1 to 0 with probabilities on the Y axis and the number of measures on the X axis. This graph shows effect sizes of .5 or less should yield non-significance with sample sizes below 120 measures. The other graphs show results with as many as 10 small sample size replications. There is a convergence of means with the effect size as sample size increases and measurement accuracy emerges.     

Does the feedback from previous trials influence current decisions? A study on the role of feedback processing in making decisions under explicit risk conditions

Decision situations frequently provide information about the amount of gains and losses and winning probabilities. In decisions under these conditions, also called risk conditions, both the use of feedback and executive functions have been shown to influence the decision‐making process, as revealed in different patient populations. However, the influence of offering feedback in tasks examining decisions under risk conditions has not been investigated experimentally, so far. This was the aim of the present study. For this purpose, a sample of healthy individuals was examined with the Game of Dice Task, a decision‐making task that explicitly provides the rules for gains and losses and in which participants receive feedback after each trial. In addition, a modified version of this task was performed, in which the feedback after each trial and all associated feedback components were removed. Results indicate that participants had a lower performance in the modified Dice Task without feedback. They selected the disadvantageous alternatives more frequently, when they did not receive feedback following their choices. Task performance in either version was correlated with executive functioning. Conclusion: In decisions under risk conditions, both executive functions as well as the use of feedback following previous trials are important components for optimal performance. Results have implications for the interpretation of deficient decision making in patients with neuropsychological impairments as both disturbances in categorization and other cognitive processes as well as emotional dysfunctions can compromise decision making in risky situations.     

Gabaergic mechanisms of hypothalamic nuclei in the expression of conditioned fear

The amygdala, the dorsal periaqueductal gray (dPAG), and the medial hypothalamus have long been recognized to be a neural system responsible for the generation and elaboration of unconditioned fear in the brain. It is also well known that this neural substrate is under a tonic inhibitory control exerted by GABA mechanisms. However, whereas there is a growing body of evidence to suggest that the amygdala and dPAG are also able to integrate conditioned fear, it is still unclear, however, how the distinct hypothalamic nuclei participate in fear conditioning. In this work we aimed to examine the extent to which the gabaergic mechanisms of this brain region are involved in conditioned fear using the fear-potentiated startle (FPS). Muscimol, a GABA-A receptor agonist, and semicarbazide, an inhibitor of the GABA synthesizing enzyme glutamic acid decarboxylase (GAD), were used as an enhancer and inhibitor of the GABA mechanisms, respectively. Muscimol and semicarbazide were injected into the anterior hypothalamus (AHN), the dorsomedial part of the ventromedial nucleus (VMHDM), the dorsomedial (DMH) or the dorsal premammillary (PMD) nuclei of male Wistar rats before test sessions of the fear conditioning paradigm. The injections into the DMH and PMD did not produce any significant effects on FPS. On the other hand, muscimol injections into the AHN and VMHDM caused significant reduction in FPS. These results indicate that injections of muscimol and semicarbazide into the DMH and PMD fail to change the FPS, whereas the enhancement of the GABA transmission in the AHN and VMHDM produces a reduction of the conditioned fear responses. On the other hand, the inhibition of this transmission led to an increase of this conditioned response in the AHN. Thus, whereas DMH and PMD are known to be part of the caudal-most region of the medial hypothalamic defensive system, which integrates unconditioned fear, systems mediating conditioned fear select the AHN and VMHDM nuclei that belong to the rostral-most portion of the hypothalamic defense area. Thus, distinct subsets of neurons in the hypothalamus could mediate different aspects of the defensive responses.     

Syllable onsets are perceptual reading units

Syllable onsets are defined as the initial consonant or consonant cluster in a syllable (e.g., BR in BREAD). In the present study, using a letter detection paradigm and French words, we tested whether syllable onsets are processed as units by the reading system. In Experiment 1, we replicated Gross, Treiman, and Inman's (2000) result of observing no difference between the detection latencies of letters embedded in a multi-letter syllable onset (e.g., c in ECLATER) relative to a single-letter syllable onset (e.g., C in ECARTER). In Experiment 2, participants took longer to detect the target letter when it was in the second position of a multi-letter onset (e.g., L in TABLIER) than when it was a single-letter onset (e.g., L in ECOLIER). In Experiment 3, this position effect was replicated for graphemes. In Experiment 4, we tested and ruled out an alternative explanation of the unitization effect, in terms of lateral masking. These results indicate that syllable onsets are processed as perceptual units by the reading system and that the unitization effect is position dependent.     

The Use of Acceptance and Commitment Therapy to Promote Mental Health and School Engagement in University Students: A Multisite Randomized Controlled Trial

The aim of this multisite randomized controlled trial was to determine whether an intervention based on Acceptance and Commitment Therapy (ACT) was efficacious in improving university students’ psychological flexibility, mental health, and school engagement. Students were recruited in four Canadian universities and randomly assigned to an intervention (n = 72) or a wait-list control group (n = 72). Students in the intervention group took part in four 2.5-hour workshops during a 4-week period and were asked to do exercises at home (e.g., meditation, observation grids). Wait-list students received the intervention soon after the post measurements. MANCOVAs and ANCOVAs revealed that students in the intervention group showed greater psychological flexibility at postintervention than those in the control group. They also reported greater well-being and school engagement, and lower stress, anxiety, and depression symptoms. Taken together, results of this study suggest that an ACT-based intervention offers a valuable way to promote mental health and school engagement in postsecondary settings.     

Parental involvement in home visiting: Interpersonal predictors and correlates

Parents' commitment to and usage of early intervention are key variables in understanding discrepancies in families' susceptibility to these services. Although the important role of fathers in infant development is widely recognized, early interpersonal predictors of paternal involvement in home‐visiting programs have been understudied. This article aims to fill this gap by regressing parents' postnatal involvement on prenatal partnership satisfaction and quality of the helping relationship in a sample of 124 socially and financially disadvantaged families. Paternal program engagement was predicted by partnership satisfaction whereas the perceived quality of the helping relationship best predicted maternal program engagement, with demographical characteristics controlled. Maternal program engagement also mediated the relationship between partnership satisfaction and paternal program engagement. The results are discussed against the theoretical background.     

Examining genetic and environmental effects on reactive versus proactive aggression

This study compared the contribution of genes and environment to teacher-rated reactive and proactive aggression in 6-year-old twin pairs (172 pairs: 55 monozygotic girls, 48 monozygotic boys, 33 dizygotic girls, 36 dizygotic boys). Genetic effects accounted for 39% of the variance of reactive aggression and for 41% of the variance of proactive aggression. The remainder of the variance was explained by unique environmental effects. Genetic as well as unique environmental effects were significantly correlated across reactive and proactive aggression (genetic correlation = .87, environmental correlation = .34), but this overlap was largely due to a common underlying form of aggression (i.e., teacher-rated physical aggression). Once common etiological factors due to physical aggression were accounted for, reactive and proactive aggression shared no other genes and only a few environmental influences, although additional specific genetic and environmental effects were observed for both reactive and proactive aggression. These specific effects indicate that both reactive and proactive aggression may be influenced mostly by socialization experiences that are specific to each type of aggression and only to a very small degree by specific genes.