2017 Data Blitz Presentations

​

Benedek Kurdi: Evaluative Pairings Shift Implicit Attitudes More Durably than Evaluative Statements

​

Previous work (Kurdi & Banaji, 2017) revealed a surprising result: repeated evaluative pairings (REP; exposure to pairings of category members with positive and negative images) were significantly less effective in shifting implicit (IAT) attitudes than evaluative statements (ES; merely informing participants of upcoming stimulus pairings without actual exposure). The combination of ES + REP did not outperform ES, suggesting that this condition reflects learning from ES alone, without REP producing any added benefit. The present studies probed the boundary conditions of this finding by exploring the stability of REP-based vs. ES-based evaluative learning over time. Study 1 (N = 4217) used a moving average method to investigate how implicit attitude strength evolved during an IAT session administered immediately after learning. Whereas implicit attitudes created via REP remained stable, those created via ES or ES + REP decayed rapidly. In Study 2 (N = 254), when implicit attitudes were measured following a 20-minute delay, REP and ES + REP both led to descriptively greater learning effects than ES in isolation. ES + REP was no more effective than REP alone, suggesting that unlike in earlier work, ES had no incremental effect above and beyond REP. Evaluative learning produced via ES is superior immediately, but REP appears to have greater staying power over time.

 

​

Tatiana Lau: Overcorrection for Social-Categorization Information Moderates Impact Bias in Affective Forecasting

​

Plural societies require us to forecast how in-group and out-group members will respond to gains and setbacks. Typically, correcting affective forecasts to include more relevant information improves accuracy by reducing their extremity. In contrast, we find across five experiments that providing forecasters with social category information about targets makes forecasts more extreme and less accurate. In both political (winning and losing the Midterm Elections, N = 1044; losing an online tournament, N = 512) and sports contexts (losing the Harvard-Yale football game, N = 350), forecasters exhibited greater impact bias for both in-group and out-group members (e.g., a "Democrat" or "Republican"; effect sizes = 14.30, 11.78, 7.82, 3.22 for in-group, 19.70, 14.75, 17.82, 4.36 for out-group) than for unspecified targets (a "person"; effect sizes = -1.42, -0.74, 2.63, 1.05) when predicting their responses to events. Inducing time pressure (N = 1445) reduced the extremity of forecasts for group-labeled (effect sizes = 8.63, 13.81) but not unspecified targets (effect size = 3.91), suggesting that increased impact bias was due to overcorrection for social category information. Finally, overcorrection was better accounted for by stereotypes than retrieval of extreme exemplars; 545 participants ranked targets as average in terms of affective extremity (mean differences = 3.86, 11.58 for in-group and out-group compared to unspecified). Because many of the decisions that we make for others rely on these affective forecasts, the insight provided here may help to reduce bias in many domains of decision-making.

​

 

Katie Hoemann: A network theoretic approach to emotional granularity

​

Emotional granularity describes the quality of a person‰Ûªs emotional experience, and the ability to differentiate aspects of this experience via language (Barrett, Gross, Christensen, & Benvenuto, 2001). Highly granular individuals experience emotions that are finely tuned to a given event, and consistently mark the difference between various experiences with words. Emotional granularity has been linked to a host of positive mental and physical health indicators (Kashdan, Barrett, & McKnight, 2015), including less alcohol consumption when stressed (Kashdan, Ferssizidis, Collins, & Muraven, 2010). Measurement of emotional granularity relies on experience sampling protocols, in which participants rate the intensity of their emotional experience on a variety of pre-populated terms (e.g., angry, disappointed, surprised). Although participants are typically sampled at multiple time points per day, across multiple days, to date these online ratings have only been analyzed in aggregate using correlational approaches (Tugade, Fredrickson, & Barrett, 2004) ‰ÛÒ a static snapshot of patterns in term co-endorsement that fails to capture individual variation over time. Here, we use network analysis (Boccaletti, Latora, Moreno, Chavez & Hwang, 2006) as an analytic framework for modeling emotional granularity, including dynamics over time. We explore how network analyses can be used to describe individual differences in the semantic space for emotion concepts (building on work by Toivonen et al., 2012), and extend this approach to how network metrics (e.g., number of subgraphs, network coherence) may be used to operationalize other aspects of emotional experience, such as stability versus variation in patterns over time.

 

Kristina Wright: Central Gray Neurons Rapidly Signal Threat Probability

​

The ability to discriminate between a range of safe and dangerous stimuli is vital. Further, the rapidity and accuracy of fear responses informed by this discrimination are imperative. Fear conditioning procedures, combined with single unit recording, allow for investigation of neural activity relevant to appropriate discrimination across a range of threatening stimuli. Using a fear conditioning procedure, we examined the relationship between neural activity in the central gray, an area highly relevant to fear-output, and discrimination of threat. Adult, male, Long Evans rats were implanted with drivable, microelectrode bundles aimed at the central gray. Following recovery, rats were presented with three auditory cues, each associated with a different probability of foot shock: safety, p = 0.00; uncertainty, p = 0.38; and danger, p = 1.00. All rats demonstrated excellent fear discrimination: achieving high fear to the danger cue, little or no fear to the safety cue, and intermediate fear to the uncertainty cue. Our analysis of central gray activity during discrimination revealed a large population of neurons demonstrating both rapid and phasic increases in firing at the onset of one or all three cues. Interestingly, the magnitude of increased firing at cue onset was best explained by the probability with which the corresponding cue predicted foot shock. This neural representation of precise, rapid fear discrimination, preceding a behavioral response, may play a critical role in the rapid and accurate organization of an appropriate fear response.

 

​

Chris Lim: Unpredictable Aversive Events Enhance Fear Memory Formation

​

Classical auditory conditioning is commonly used to study the strength of fear memory in post-traumatic stress disorder (PTSD). However, traumatic event does not happen in a predictable manner in real life. To study the effect of the unpredictable traumatic event on fear memory formation, we modified the standard auditory conditioning to resemble the occurrence of the unpredictable traumatic event in real life. Long-Evans rats were exposed to either the predictable auditory fear conditioning that consisted of auditory cues (conditioned stimulus, CS) followed by footshocks (unconditioned stimulus, US) at a fixed CS-US onset interval (interstimulus interval, ISI), or the unpredictable auditory fear conditioning with variable ISI. Fear memory strength was then evaluated by analyzing the freezing time within the tone during the tone extinction. Our results showed that the unpredictable fear conditioned rats exhibited significantly more freezing compared with the predictable fear conditioned rats. Alternatively, unpredictable auditory fear conditioning could also be performed by delivering the footshock at the termination of the tones that had variable duration. This paradigm made the tone uninformative to predict the footshock. Rats that were exposed to this paradigm also displayed significantly higher freezing behavior while comparing with the predictable fear conditioned rats. Furthermore, to eliminate the effect of the shortest ISI on the fear memory formation, we exposed the rats to either the predictable or the unpredictable conditioning in which both carried the same minimum ISI. Still, the unpredictable conditioning significantly elevated the rats freezing behavior. Overall, we demonstrated that unpredictable aversive events enhanced the fear memory formation. Our results implicate that our modified fear conditioning is a more appropriate paradigm that could replicate the real life traumatic experience in laboratory settings. 

 

Hilary Richardson: The Development of Brain Regions for Thinking about the Internal States of Others

​

In human adults, distinct networks of brain regions are recruited to think about the bodies (physical sensations) and minds (mental states) of others. The current study characterizes the developmental trajectory of these two functionally dissociable networks, and tests for relationships between functional change in these networks and behavioral developments in reasoning about the minds of others (“theory of mind”, ToM). A large cross-sectional sample of children ages three to twelve years old (n=122), in addition to a group of adults (n=33), watched a short, animated movie while undergoing fMRI. The movie highlights the physical sensations (often pain) and mental states (beliefs, desires, emotions) of the main characters, and, critically, provides an experimental context that is feasible and engaging for even the youngest children. We conduct interregional correlation analyses in order to characterize the extent to which ToM and pain brain networks are functionally distinct, and reverse correlation analyses in order to determine which events in the movie evoke responses in each network, and to characterize functional maturity of the two networks in children. We find evidence that 1) ToM and pain networks are functionally distinct by age three years, 2) functional specialization increases throughout childhood, and 3) functional maturity of each network is related to increasingly anti- correlated responses between the two networks. Further, these data provide evidence that the most studied milestone in ToM behavioral development, passing explicit false belief tasks, does not correspond to dramatic changes in the neural basis for reasoning about the minds of others. 

​

​

Poster presentation abstracts and presenters coming soon 

​