Fall Academy 2018

Internal Information

Date: October 12-15, 2018
Location: University of Virginia, Charlottesville
Organizer: Steven M. Boker

LECTURES

Please find the PDFs of the readings at the bottom of this box!

Bifactor models for predicting criteria by general and specific factors: Problems and alternatives
Michael Eid, Freie Universität Berlin

• Eid, M., Geiser, C., Koch, T., & Heene, M. (2017). Anomalous results in g-factor models: Explanations and alternatives. Psychological Methods, 22, 541–562. https://doi.org/10.1037/met0000083
• Eid, M., Krumm, S., Koch, T., & Schulze, J. (2018). Bifactor models for predicting criteria by general and specific factors: Problems of nonidentifiability and alternative solutions. Journal of Intelligence, 6(3): 42. https://doi.org/10.3390/jintelligence6030042

The role of self-concept of ability in understanding achievement across development
Pamela Davis-Kean, University of Michigan

• Susperreguy, M. I., Davis-Kean, P. E., Duckworth, K., & Chen, M. (2017). Self-concept predicts academic achievement across levels of the achievement distribution: Domain specificity for math and reading. Child Development. Advance online publication. https://doi.org/10.1111/cdev.12924
• Watts, T. W., Duncan, G. J., Chen, M., Claessens, A., Davis-Kean, P. E., Duckworth, K., ... Susperreguy, M. I. (2015). The role of mediators in the development of longitudinal mathematics achievement associations. Child Development, 86(6), 1892–1907. https://doi.org/10.1111/cdev.12416

Estimation, replicability and falsifiability of latent factors using exploratory graph analysis
Hudson Golino, University of Virginia

• Golino, H. F., & Demetriou, A. (2017). Estimating the dimensionality of intelligence like data using Exploratory Graph Analysis. Intelligence 62, 54–70. https://doi.org/10.1016/j.intell.2017.02.007
• Golino, H. F., & Epskamp, S. (2017). Exploratory graph analysis: a new approach for estimating the number of dimensions in psychological research. PlosOne 12(6): e0174035. https://doi.org/10.1371/journal.pone.0174035

Shifting incentives from getting it published to getting it right
Brian Nosek, University of Virginia

• Camerer, C. F., Dreber, A., Holzmeister, F., Ho, T.-H., Huber, J., ... Nosek, B. A., ... Wu, H. (2018). Evaluating the replicability of social science experiments in Nature and Science between 2010 and 2015. Nature Human Behaviour, 2, 637–644. https://doi.org/10.1038/s41562-018-0399-z
• Munafo, M. R., Nosek, B. A., Bishop, D. V. M., Button, K. S., Chambers, C. D., Percie du Sert, N., ... Ioannidis, J. P. A. (2017) A manifesto for reproducible science. Nature Human Behaviour, 1: 0021. https://doi.org/10.1038/s41562-016-0021

Tracking neuro capacity with an experiment and modeling ecosystem
Per B. Sederberg, University of Virginia

• Eid et al., 2017 538.46 kB
• Eid et al., 2018 4.66 MB
• Susperreguy et al., 2017 192.66 kB
• Watts et al., 2015 159.31 kB
• Golino & Demetriou, 2017 1.37 MB
• Golino & Epskamp, 2017 10.17 MB
• Camerer et al., 2018 1.22 MB
• Munafo et al., 2017 396.32 kB
• Gravina & Sederberg, 2017 586.53 kB
• Nielson et al., 2015 1.15 MB
• Vaish, 2018 201.75 kB
• Vaish et al., 2018 1.43 MB

Fellows' Presentations - Abstracts

Adjustment disorder in cardiovascular disease patients: An intensive-longitudinal study
Tania Bermudez, UZH
(Advisors: Urte Scholz, Andreas Maercker)

Cardiovascular disease (CVD) is the leading cause of death worldwide. There is a high prevalence of psychopathology among CVD patients. This includes adjustment disorder (AjD), which has been neglected in research until recently due to an ambiguous definition. This definition has led to unreliable prevalence estimates and frequent misdiagnoses. The recently completed revision of the International Classification of Diseases (ICD-11) includes a more specific definition of AjD that allows more precise diagnosis and research. Therefore, the goal of the present project was to analyze the prevalence of AjD among CVD patients, as well as the development of AjD symptoms during and right after cardiac rehabilitation on a daily basis. The study has a multicentrered, observational, intensive-longitudinal design, and is currently in the recruitment and data collection phase. Preliminary descriptive results regarding the prevalence of AjD and its development will be presented. Results should increase awareness regarding AjD in CVD, and ultimately provide information for more successful interventions regarding psychological well-being in CVD patients.

Robot-assisted therapy in chronic stroke patients: A feasibility study looking at adherence to at-home treatment recommendations
Walter Bierbauer, UZH
(Advisor: Urte Scholz)

Stroke patients often face reduced mobility and loss of upper extremity function. Robot-assisted therapy can help patients significantly improve functioning in their paretic arm. It remains unclear whether these improvements also translate into everyday life. This study investigates how stroke patients adhere to at-home treatment recommendations.
In cumulative N-of-1 trials adopting an A-B-C-B-C experimental design three patients got both solo robot-assisted therapy (B) and dyadic robot-assisted therapy (C). At-home arm activity was measured by wrist-worn accelerometers and motivational, volitional and social variables were gathered in self-reports across five weeks on a daily basis.
In weeks with robot-assisted therapy, one patient increased the participation in activities of daily living. Additionally, dyadic therapy was associated with increased paretic arm involvement in everyday life, thereby following treatment recommendations. Two patients showed no transfer effects of robot-assisted therapy to everyday life paretic arm usage.
Fostering daily at-home exercises of a paretic arm is an important goal in stroke rehabilitation. By examining the effects of different robot-assisted therapies on exercising behavior, this study contributes to the optimization of the stroke rehabilitation process. N-of-1 trials can evaluate treatment effectiveness in a single individual, thereby helping to close the gap between evidence and practice. The aggregation of trial findings could inform treatment decisions for other patients unaffiliated with the trials.

Influence of model competency on memory and attention during word learning
Julia Brehm, UZH
(Advisors: Anja Gampe, Moritz M. Daum)

Already at 14 months, children selectively learn from competent over incompetent models. Little is known about cognitive mechanisms behind this phenomenon. Differences in learning from competent and incompetent models could be the result of differential processing during phases of the learning process.
Two important cognitive mechanisms involved in learning are attention and memory. Both can act differentially on sequential stages of selective learning: (1) while experiencing a model’s competency, (2) during novel learning instances and (3) during memory consolidation, and (4) during recall of learned information. Besides memory and attention, another process might influence whether we can observe learning: Even if children learn novel information from incompetent models, they might actively suppress their knowledge and not reproduce it due to uncertainty in the correctness of the information.
In my talk, I will discuss on a theoretical level how and in which of the phases these processes could contribute to the selective learning phenomenon. Then, I will present eye-tracking data from 45 2-year-old children tested in a classical selective learning paradigm to explore attentional processes involved in selective learning. We analyzed attention allocation during selective learning in a dynamic framework applying recurrence quantification analysis (RQA). We found no differences in the dynamics of visual attention between groups in any of the phases. These results suggest that attention has no or little involvement in observed learning outcomes in selective learning and calls for a greater focus on memory processes in future studies.

Towards modeling change in cognitive processes using the COGITO study data
Tiago Cabaço, HU
(Advisor: Manuel Völkle)

Over the years, the study of response times has been dominated by the use of cognitive models. Cognitive models are statistical models that decompose observed performance into latent psychological processes. When observations from multiple individuals are available, the use of cognitive models allows the investigation of how individual differences in performance relate to differences in psychological processes. Moreover, when the same individuals are measured over multiple time points, one can investigate how changes in performance over time might be related with the temporal dynamics of cognitive processes. For this purpose, we analyzed the performance of 204 subjects (101 younger and 103 older adults) over an average period of 101 measurement occasions on two-choice reaction time tasks (CRT) from the intensive longitudinal phase of the COGITO study. We modeled cognitive processes using the drift diffusion model (DDM), which accounts for the reaction time and accuracy distributions from CRT. In this talk, we will explore how we can proceed to model the dynamics of change in the parameters of the DDM and its relationship with behavioral performance. We will address different modeling approaches that reflect alternative hypothesis for how response processes change over time. Finally, we will address potentialities and limitations of jointly modeling cognitive processes, individual differences, and dynamical processes.

Age-differential effects of opportunity costs on exhaustion and recovery
Brian B. Cardini, UZH
(Advisor: Alexandra M. Freund)

How do we know when it is time to disengage from an exhausting or relaxing activity? In this talk, I will first present a motivational account of exhaustion and recovery, according to which (i) increasing opportunity costs, (ii) decreasing mood, and (iii) expanding subjective time perception indicate to a person that the current (effortful or relaxing) activity is producing more costs than benefits and should thus be discontinued. I will then present a micro-longitudinal study that investigated age-related differences in the psychological indicators of exhaustion and recovery. Specifically, we tested the hypothesis that older adults are more sensitive to increasing opportunity costs during an effortful (relaxing) activity and therefore feel exhausted (recovered) faster than younger adults. N = 147 participants (n = 49 young adults, 48 middle-aged adults, and 50 older adults) were first exhausted by a 20-minute high intensity interval training (exhaustion period). Immediately afterwards, they listened to a 20-minute mindfulness-based relaxation video (recovery period). During the exhaustion and recovery period, participants reported their momentary states of physical and mental recovery, mood, time perception, opportunity costs, and boredom (10 measurement occasions during the exhaustion period and 10 measurement occasions during the recovery period). Older adults reported a faster increase in physical exhaustion and opportunity costs during the workout compared to younger adults. However, older adults did not differ from younger adults in their reported physical recovery and opportunity costs during the relaxation video. These results paint a nuanced picture of the relationship between opportunity costs, exhaustion, and recovery.

Age differences in the contextualization of memories
Anna Karlsson, MPIB
(Advisor: Myriam Sander)

Episodic memory involves the formation and retrieval of detailed memories, made possible through the binding of item and context information. It is well established that young adults (YA) outperform older adults (OA) in most episodic memory tasks. However, age differences are more pronounced in memory for context compared to memory for content and in associative compared to item memory. Nonetheless, recent studies have shown that OA can use contextual information to boost memory performance to the same extent as YA while at the same time showing reduced explicit memory for these contexts and for the specific item-context associations. Most studies so far are purely behavioral and thus have overlooked the neural underpinnings of these effects, whereby it remains unknown if such age differences are due to processes at encoding or retrieval.
In the present study, we use EEG and fMRI to investigate age differences in the neural substrates of episodic memory formation. Specifically, we aim to demonstrate how the quality of the item-context integration process during memory formation mediates changes in the item-specific neural memory representation, reflecting successful binding of item and context, and predicts the influence of contextual information during memory retrieval.

Factor score estimation in multi-method planned missing data designs
Mario Lawes, FU
(Advisor: Michael Eid)

Multi-method planned missing data designs such as the two-method measurement design (2-MM-PMD; Graham, Taylor, Olchowski, & Cumsille, 2006) or the three-method measurement design (3-MM-PMD; Lawes, Schultze, & Eid, 2018) are an elegant way to cost-efficiently incorporate expensive gold standard methods as well as cheaper but biased methods in research designs. In my talk, I am going to present the basic rationale of multi-method planned missing data designs as well as results from a simulation study investigating the statistical properties of 2-MM-PMD and 3-MM-PMD designs in detail. Several possible extensions in the field of multi-method planned missing data designs, such as the questions whether these designs can be used to validly and cost-efficiently estimate individual scores (i.e., factor scores), will be discussed. Examples of situations in which individual scores in such research designs would be of interest will be given. Further, a simulation paradigm to investigate the performance of different factor score estimation methods in this context will be outlined.

Is it about the "how" or "why" of goal pursuit? A study on goal focus across the entire lifespan
Lea Mörsdorf, UZH
(Advisors: Alexandra M. Freund, Moritz M. Daum)

Setting and pursuing goals guide our behavior and accompany us throughout life. We refer to goals as cognitive representations of means-ends associations, that is, the link of means that lead to a desired state. Here, we ask how the relative salience of means and ends changes across the lifespan. Previous research on adulthood suggests that, while younger adults tend to focus on the outcomes, older adults concentrate more on the processes of goal pursuit. As of yet, there is no systematic research into the development of goal focus across the entire lifespan, including childhood and adolescence with comparable paradigms. In this study, we investigated goal focus with a sample (as of yet) of N = 307 participants between 3 and 85 years who were tested in individual sessions in our laboratory. Following a multimethodological approach, we used non-verbal (e.g., eye tracking, imitation choice task) as well as verbal paradigms (e.g., thinking exercise, action descriptions) to investigate whether they concentrate more on the processes or outcomes. In contrast to the previous research on adult changes in goal focus, we found only very little evidence for age-related differences: Only in one task, the action descriptions either in terms of means or ends, we found evidence for age-related differences, and those were in the opposite direction than expected. With increasing age, participants chose more statements describing an outcome (estimate = 0.02, SE < 0.01, p < .001). No other tasks revealed age-related differences in goal focus. These results and next steps will be discussed.

Tangle: Defining a new measure of time series complexity
Robert Moulder, UVA
(Advisor: Steven M. Boker)

Time series are a rich source of information for psychological, biological, and behavioral researchers interested in understanding differences between groups of participants. A common method for analyzing between-group time series analysis is to compare group distributions of simple statistics derived from each participant’s time series data (e.g., mean and variance). While informative, these statistics fail to capture many of the dynamic and complex properties of time series data. However, many current quantifications of time series complexity require a hindering level of technical ability, minimum sample size for reliable estimation, or amount of computational power in order for researchers to be able to use such quantifications. In this presentation, I propose a technically simple iterative method for quantifying complexity in short time series derived from computer science research. Results of a simulation show this quantification remains reasonably stable for time series containing at least 20 time points.

Trajectories of family instability and disruptive behaviors across early childhood: A longitudinal prospective study of at-risk families
Sean Womack, UVA
(Advisor: Mel Wilson)

Economically marginalized families are at a particularly high risk to experience instability in the form of residential mobility, family structure instability, and incarceration of parenting figures. Previous research has linked instability in early childhood to later behavior problems, but little is known about the longitudinal relationship between instability and behavior problems. The present study uses data from 731 families recruited to be at high risk for child behavior problems on the basis of socioeconomic disadvantage, family problems, and child behaviors. Parents reported on residential mobility, family structure instability, incarceration of adults in the home, and child behavior problems at child age 2, 3, 4, 5, and 7.5. Growth curve models were fit to model the growth of instability, disruptive behaviors, and the covariance between the instability and disruptive behavior growth factors. Both instability and disruptive behaviors were found to decline over time. Instability between birth and 2 predicted higher levels of disruptive behaviors at age 2 and the change in instability over time positively predicted the change in disruptive behaviors over time. The findings from the present study have implications for mental health professionals seeking to intervene on behavior problems at the family level as well as policy makers who are making decisions at the community level.

Fellows' Posters: Abstracts

Applying reinforcement learning to understand personalized, contextual effectiveness of emotion regulation strategies
Miranda Beltzer, UVA
(Advisor: Bethany Teachman)

Emotion dysregulation, or the ineffective use of strategies for managing one’s own emotions and their intensity, has been found to play an important role in many mental disorders (Fernandez, Jazaieri, & Gross, 2016). Although theories about emotion dysregulation have long recognized the importance of using the right emotion regulation strategy for a given situation, practically, researchers have tended to categorize strategies as generally more adaptive or maladaptive (Bonanno & Burton, 2013). More recently, however, researchers have begun to examine how the effectiveness of a given strategy may vary based on personal factors, such as gender, age and depression severity, and contextual factors, such as current experienced emotion, intensity of emotion, and time of day (Aldao & Nolen-Hoeksema, 2012; Nolen-Hoeksema & Aldao, 2011). In this ongoing study, we aim to leverage big data to create an algorithm that recommends emotion regulation strategies likely to be effective for a given individual in their current context. N = 125 high social anxiety adults will complete five weeks of experience sampling through their personal smartphones, including six surveys throughout each day assessing current affect, anxiety, social context, and recent attempts at emotion regulation. We will also gather data from passive smartphone sensors, including GPS. We will use reinforcement learning to understand the contextual effectiveness of each emotion regulation strategy, and then will cluster individuals into groups with similar patterns of contextual effectiveness. We will then use features measured at the baseline lab session (including demographic and psychological variables) to predict cluster membership.

God has a bigger plan for people without friends: Religion as purpose for the socially disconnected
Todd Chan, UM
(Advisor: Oscar Ybarra)

Social relationships are associated with purpose in life. People who are socially disconnected report a lowered sense of purpose. How can these people compensate for purpose in life? We propose that religious beliefs uniquely compensate for the purpose in life by providing (a) broader purpose to turn to, and (b) a relationship with God that substitutes for human relationships. In three studies, we analyze nationally representative and longitudinal datasets using moderated regression and cross-lagged panel analyses. We find that religious beliefs confer minimal additional purpose for socially connected individuals; however, for socially disconnected individuals, increasing religiosity predicted higher levels of purpose in life. Results suggest that although people primarily derive purpose from social relationships, people who lack these relationships may be able to leverage their religious beliefs for purpose until they can reconnect.

Mapping sentences to causal events in Turkish and Swiss-German
Ebru Ger, UZH
(Advisor: Moritz Daum)

Children use syntactic frame (number of noun phrases) as a guide to map novel verbs to causal meanings (Fisher et al., 2010). However, most previous research has been conducted in English where the prominent cue to causality is the syntactic frame. Other languages which have other cues to causality such as verb morphology are understudied. In Turkish, for example, a causative marker can be affixed to verbs to express causality. Moreover, it is a pro-drop language and thus, it is possible to form sentences containing only verbal morphological but no syntactic cue to causality. In this study, we examined how Turkish and Swiss-German speaking children perform in mapping causal sentences to causal scenes. Each of three Turkish-speaking groups hear sentences that contain (i) only the verbal morphological cue, (ii) only the syntactic cue, or (iii) both cues. We compare each group’s performance to a Swiss-German-speaking group who hear sentences with only the syntactic cue, as Swiss-German lacks a verbal morphology cue. We use a preferential pointing task adapted from Kline et al. (2017), where children point to either a causal scene or a non-causal scene upon hearing causal sentences involving nonsense verbs. Hypotheses are discussed under the framework of a universalist versus emergenist approach (Lidz et al., 2003).

Making use of the longitudinal nature of an event-related EEG design to investigate tinnitus
Laura Jagoda, UZH
(Advisor: Martin Meyer)

Subjective tinnitus is the perception of sound in the absence of an external sound source. It represents a widespread symptom with reported prevalence rates varying between 5% and 43% worldwide. While most individuals affected learn to adapt or compensate, a small percentage continues to suffer significantly. According to a new framework, based on predictive coding, tinnitus becomes chronic when perceptual inference mechanisms learn to expect tinnitus, overriding the default percept of silence with that of tinnitus. Excessive focused attention and affective reactions (via limbic connections) can further amplify the expectation of the tinnitus percept. So far neurophysiological research on tinnitus is primarily focused on resting state measurements, investigating spontaneous oscillatory alterations or network changes in tinnitus individuals (TI). When analyzing event-related EEG data, responses are usually averaged over the presented number of trials to achieve a suitable signal-to-noise ratio, as single neural responses to sensory stimuli are very small. This results in a loss of information about changes in event-related responses over the course of the experiment. Since the adaption to sensory input over time plays a key role in the chronification of tinnitus, a moving-window single-trial approach (Hasenstab et al., 2015) was used to analyze event-related responses of TI in an auditory oddball paradigm. The traditional analysis showed that TI with higher distress reacted more strongly (higher N2 peaks) to the irregularly presented salient stimuli. The single-trial analysis additionally revealed that TI with higher distress seem to habituate faster to the regularly presented standard stimuli. Hence, the hyperresponsiveness to the salient stimuli might in fact be explained by an increased sensory precision resulting from an excessively fast formation of predictions.

Spotlight on the sleeping brain! The role of sleep in memory consolidation across childhood
Ann-Kathrin Joechner, MPIB
(Advisor: Markus Werkle-Bergner)

Recently acquired memories are believed to be transferred during Non-Rapid Eye Movement (NREM) sleep from the hippocampus to the neocortex and integrated into long-term storage. Functionally, this consolidation process has been associated with an interplay between cortical slow oscillations (SO), thalamo-cortical sleep spindles (Sp), and hippocampal ripples in adults. Maturation is characterized by an increase in general cognitive abilities and mnemonic functions in parallel with quantitative and qualitative changes in sleep parameters and brain structure. However, the role of sleep in memory consolidation and the constitution of sleep’s cardinal oscillations across maturation is still elusive. Therefore, my dissertation project aims at the characterization of (1) SOs and Sps in children of different age and their interrelation with brain structure and (2) the association between these parameters with the maturation of memory processes. For this purpose, we conducted a pilot study to examine sleep-dependent evolution of associative memories of different quality and their relation with Sps in 5- to 6-year-old children.

Learning: Effectiveness or memory?
Angela Jones, MPIB
(Advisor: Azzurra Ruggeri)

What do we prioritize when learning something new, for example, how to activate a novel toy? Being efficient, that is, finding the answer as quickly as possible, or maximizing long-term memory retention? In this study, we explore the impact of different question-asking strategies on task-related memory across the lifespan. Five-, seven-, and ten-year-olds and adults will play three rounds of the 20-Questions game, where they have to identify, amongst a set of objects, the one that activates a novel toy by asking as few yes/no questions as possible. We will assess their memory for the target object, the other objects presented, the specific toy that object activated, and the questions they had asked, at three time points. Previous literature shows that question-asking efficiency increases steadily across the lifespan. Younger children’s inefficiency is generally considered a disadvantage, but it might confer other benefits: less efficient strategies may result in better long-term memory for some of the stimuli considered, the relationships learned, or the learning process. In this sense, young children’s strategies, despite being less efficient, might help them enrich and strengthen their knowledge base to provide a strong foundation for cognitive skills such as categorization and some components of metacognition.

Fighting fair with parents and peers: Differential impacts of conflict negotiation processes on adolescent sexual and romantic relationship experiences
Jessica Kansky, UVA
(Advisor: Joe Allen)

In early adolescence, the developmental challenge of seeking autonomy from parents and increasing integration with same-aged friends becomes dominant (Steinberg, 2001) before individuals are tasked with navigating romantic relationships by late adolescence. Evidence suggests a cascade effect in relationship quality with parents to peers to romantic partners (Oudekerk et al., 2015) and points to deviant peers as a predictor of poor early romantic relationships (Capaldi et al., 2004). However, it is less clear whether healthy parent and peer relationships that simultaneously promote autonomy and connection impact romantic and sexual involvement differently. This study addresses how behaviors that promote autonomy and relatedness with peers and parents relate to sexual and romantic involvement and quality.
Multi-informant longitudinal data were obtained for a diverse (race/ethnicity, gender, and income) community sample of 184 adolescents (86 males, 98 females) along with their mother, father, and close friend at age 13. Behaviors promoting autonomy and relatedness within each relationship (i.e, teen-peer, teen-mom, teen-dad) were assessed at age 13 using an 8-minute, video-taped conflict task. Observed positive interactions include responding confidently, rationally, clearly, and in an engaged manner. Follow-up data was collected at age 17 via self-report questionnaires regarding sexual history (age of first sex, number of relationships and sexual partners) and current romantic relationship (positive and negative conflict management).
Regression analyses suggest positive interactions within teen-peer pairs at age 13 was associated with delayed age of first sexual experience and fewer sexual partners and relationships by age 17. Delayed sexual initiation is associated with less risky behaviors (Coker et al., 1994) indicating that healthy peer relationships may buffer against the detrimental effects associated with earlier sexual and romantic involvement. Positive interactions within teen-mother relationships at age 13 were related to greater conflict (positive and negative) within romantic relationships at age 17, while positive interactions within teen-father relationships were only related to greater positive conflict.
Findings suggests that teen’s relationships with parents and peers may influence their romantic trajectories in unique ways. Specifically, results indicate that peers play a role in the timing and number of relationships and sexual partners but parents play a greater role in the quality of these relationships. Taken together, this suggests that the developmental significance of autonomy for later romantic experiences can differ by type of relationship. Implications of these findings for future research and practice will be discussed.

Short-term variability of future time perspective in everyday life
Marko Katana, UZh
(Advisors: Mike Martin, Lutz Jäncke)

Despite well-documented age differences in future time perspective (FTP) across the lifespan, little is known about the extent to which FTP fluctuates within-person over a short timescale and how it is associated to short-term fluctuations of affect. This micro-longitudinal study over the course of ten days has investigated the short-term variability of FTP and the intraindividual coupling between FTP and positive as well as negative affect. The study examined a sample of 564 adults ranging from 20 to 75 years (M = 48.3, SD = 10.0). Multilevel modeling results showed that on days when individuals reported a more open-ended FTP, they also reported higher positive and lower negative affect. Moreover, results showed that older adults reported weaker within-person associations between FTP and affect compared to younger adults. Overall, these findings advance our understanding of perceptions of FTP in daily life as a time-variable construct and its association to affect across the adult lifespan.

Infants’ brain responses to pupillary changes in others are affected by race
Caroline M. Kelsey, UVA
(Advisor: Tobias Grossmann)

Sensitive responding to eye cues plays a key role during human social interactions. Recently, infants have been found to mimic observed pupillary changes in others, instantiating a foundational mechanism for eye-based social communication. Among adults, perception of pupillary changes is affected by race. Here, we examined whether and how race impacts the neural processing of others’ pupillary changes in early ontogeny. We measured 9-month-old infants’ brain responses to dilating and constricting pupils in the context of viewing own-race or other-race eyes using functional near-infrared spectroscopy (fNIRS). Our results show that only when responding to own-race eyes, infants’ brains distinguished between pupillary cues. Specifically, infants showed enhanced responses in the right superior temporal cortex when observing own-race pupil dilation. Moreover, when processing other-race pupillary changes, infants recruited the dorsolateral prefrontal cortex, a brain region linked to cognitive control functions. These findings suggest that, early in development, the fundamental process of responding to pupillary changes in others is hampered during interracial interactions and that such interactions may afford greater cognitive control or effort. This critically informs our understanding of the early origins of responding to pupillary signals in others and highlights the impact of race on the effective detection of social signals.

How age shapes neural representations:
fMRI evidence for broader tuning functions in older adults
Christoph Koch, MPIB
(Advisor: Nicolas Schuck)

Evidence from fMRI studies in humans suggests an age-related loss in distinctiveness of neural patterns (neural dedifferentiation). One of the suggested mechanisms behind this is the broadening of tuning curves on a cellular level. This phenomenon might lead to specific cells responding to a broader range of stimuli, rendering representations on a network level less distinct from one another. This assumption holds two main predictions: First, network representations of categories in continuous space should vary depending on category distance. That is, two categories neighbored in space should be represented more similar than two distant categories. Second, with increasing age, this pattern should become more pronounced. To test these predictions, we analyzed fMRI activity patterns of traveled directions in a 3D virtual environment in older and younger adults using a multivariate pattern classification approach. Analysis revealed neighbored directions were systematically more often confused with the target direction, resulting in confusion matrices in the shape of neural tuning curves with older adults showing a broader shape. These results are in line with assuming a role of lost selectivity in encoding continuous information, such as spatial direction, and may suggest a putative role of similar population level effects in memory aging more generally.

Take a break from your goals? Antecedents and consequences of goal shelving
Zita Mayer, UZH
(Advisor: Alexandra M. Freund)

Most people pursue multiple goals at the same time, both within and across different life domains. Balancing the demands of multiple goals can be challenging, as our goal pursuits often draw on finite resources (e.g., time, energy, money). To resolve resource-based goal conflicts, people may choose to shelve goals: withdraw goal investments with the intention to return. We present our hypotheses on (1) antecedents of goal shelving, (2) proximal affective and motivational consequences of goal shelving, and (3) age-related differences in preferences for and consequences of shelving vs. abandoning conflicting goals.

Bias correction for replacement samples in longitudinal research
Jessica Mazen, UVA
(Advisor: Cynthia Tong)

Missing data is a commonly encountered problem in longitudinal research. One way researchers handle missing data is through the use of supplemental samples (i.e., the addition of new participants to the original sample after missing data appear at the second or later measurement occasions). Two types of supplemental approaches are commonly used: a refreshment approach where researchers select additional participants using the same criteria as the initial participants (i.e., random selection from the population of interest) and a replacement approach where researchers identify auxiliary variables that explain missingness and select new participants based on those attributes. Past research demonstrates that the use of replacement samples results in biased parameter estimates. However, replacement samples have been used in previous studies which may have led to biased parameter estimates and misleading statistical conclusions. Furthermore, there are no studies addressing how to correct this bias. Thus, for this study, we propose three ways to correct the bias introduced by replacement samples: a parametric bootstrapping method, a non-parametric bootstrapping method, and a sample reweighing method. The three methods will be introduced in this talk and their performance will be evaluated and compared.

Factors associated with adversity-related positive development: Resilience in Irish adult survivors of childhood institutional abuse
Shauna L. Mc Gee, UZH
(Advisors: Andreas Maercker, Myriam V. Thoma)

A report (on the period from 1936 to 1999) on institutional welfare settings in Ireland, revealed that many children experienced maltreatment, neglect, and abuse, during their time in care. This included a harsh regime, childhood labor, and physical and sexual assault. Preliminary research on the adult psychological adjustment in this sample showed a high prevalence of psychological disorders in adulthood, with only a small percentage considered resilient. However, little is known about the factors which influence the development of resilience or psychopathology following such adversity.
This study therefore aims to assess adversity-related positive development (e.g., resilience, the steeling effect) in Irish survivors of institutional abuse and matched controls. It further aims to identify general, national- and culture-specific resilience and vulnerability factors related to welfare practices, which may influence health and well-being outcomes in later life.
A cross-sectional, mixed methods design is used. A quantitative questionnaire survey and qualitative semi-structured interviews are used to asses childhood adversity, stressors, resilience, coping mechanisms, and current psychological and physical health. Participants include Irish individuals who were in institutional (welfare) care in childhood/adolescence, currently aged 50 years or older, as well as matched controls.
Data collection is ongoing. Preliminary qualitative analyses (n = 7) suggest that a number of factors are emerging which are associated with adversity-related positive development. These include (1) factors during the time in institutional setting: personality traits, social support, individual coping strategies; and (2) factors during later life: personality traits, spirituality, self-motivation, others-motivation, and self-enhancement.

Lost in translation? How gender is expressed through language and how this affects audience evaluations in TED talks – A study protocol
Tabea Meier, UZH
(Advisors: Mike Martin, Andrea B. Horn)
Knowledge dissemination marks an integral part of success, and modern internet platforms such as the TED conference build new opportunities to share insights with a broad audience. In this format, individuals give short talks about their area of expertise. The videos are hosted on the TED website and have been translated into several languages by volunteers.
Previous research has identified differences in audience reactions for male and female TED speakers. In another stream of research, stable findings point at gender differences in language use, i.e., differences in the frequency with which certain type of words are used.
In the present study we will expand previous findings on language use by studying whether gender differences in language use can also be observed in the context of TED talks, and whether a gender-specific language style is transferred onto translations. In other words, we will investigate whether the original gender-specific language style of the speaker is maintained in the translated talks, even if the translator is from the opposite gender. Secondly, we will study how gender of TED speakers in combination with their language use affects audience evaluations by investigating two competing hypotheses: Does a more "masculine" language style relate to more positive talk ratings (masculinity-as-higher-status-hypothesis)? Or is it rather a gender-conform language style that leads to more positive talk ratings (gender-conform-language-hypothesis)?
N = 1,648 English TED talk transcripts along with their German translations and talk ratings have been collected from the TED website. Gender of speakers and translators was referred from the videos and personal profiles.
The results of this study will contribute to a scientific understanding of gender differences in language use, by expanding previous findings onto the context of translations. Furthermore, insights will be gained about how a speaker’s gender and language use impacts audience evaluations.

Speech intelligibility in noise: Influence of hearing abilities, cognition, and brain structure
Pia Neuschwander, UZH
(Advisor: Martin Meyer)

Older adults often show problems with speech intelligibility in noisy environments. Age-related hearing loss is an explanation for this well-known phenomenon, but it cannot be fully explained through it, as even people with assumed good hearing often show impairments of speech intelligibility in noisy environments (Giroud et al., 2018). As previous studies have shown, cognitive factors and the anatomy of auditory related brain areas also seem to be important for good speech understanding in older age. Therefore, the aim of this study is to shed more light on this complex interplay between speech intelligibility, cognitive functions, and brain structure. So far, we have analyzed data from 61 participants between 64 and 78 years of age. They all underwent four sessions of data acquisition to assess their cognitive skills (working memory, attention, inhibition, and processing speed), different hearing and speech abilities, and brain structure. Partial correlation analysis with age as a control variable and stepwise regressions with speech intelligibility performance as dependent variables have been conducted. Preliminary results will be presented.

Well-being trajectories in old age and health: The role of optimism and personality
Sophie Potter, HU
(Advisor: Denis Gerstorf)

Physical health and well-being in old age are closely associated. Even though these associations vary from person to person, inter-individual differences in subjective processes that may affect this (e.g., disposition) have often been overlooked, leaving us with a fragmented account of how health undermines well-being in old age. In the current study, we examine how morbidity and performance-based indicators of physical health relate to late-life well-being trajectories and how such associations are moderated by optimism and personality. To do so, we make use of multi-year longitudinal data obtained in the Berlin Aging Study II. We also tested whether dispositional characteristics (trait-personality and optimism) affected the strength of health–well-being associations. Growth models revealed that suffering from morbidity and poor performances on both the grip strength and perceptual speed tests were associated with lower well-being. Probing for moderation indicated that suffering from morbidity was associated with steeper well-being decline when agreeableness was high. We also found that poor performance on the perceptual speed task was associated with lower well-being when conscientiousness was low and with less favorable subsequent well-being change when openness was low. Our findings provide initial evidence for the relevance of disposition in health–well-being trajectories in old age.

Selective associations: Edibility and distaste in plant vs non-plant objects
Connair Russell, MPIB
(Advisor: Annie Wertz)

Research has shown biases in social learning towards certain evolutionarily relevant content, such as animals and danger (Barrett & Broesch, 2012). For one domain, plants, it has been hypothesized that there may be specific social learning strategies (Wertz & Wynn, 2014a), and shown that infants preferentially associate edibility with plants compared to artifacts (Wertz & Wynn, 2014b). The present work examines the associability of edibility, distaste, and neutral information with plants compared to non-plants in 18-month-olds. Participants observe an actor taking a fruit off both a plant and a feature-matched control and then, depending on condition, (i) putting the fruit to their mouth, (ii) putting it behind their ear, (iii) exhibiting distaste, (i) and (ii) being a direct replication of Wertz & Wynn (2014). Infants then choose whether they would like to eat the fruit from the plant or control. It is predicted that, in accordance with the original study, infants will preferentially select the fruit from the plant in the “in-mouth”, but not in the “behind-ear” condition. And that infants will selectively avoid the fruit from the plant in the “distaste”, but not the “behind-ear” condition. Preliminary data will be presented.

The influence of language on young children‘s understanding of causal relations
Larissa Stuber, UZH (Advisor: Moritz Daum)
Young children are faced with the challenge to unravel the causal structures in their environment. Understanding these structures is important as it enables prediction of future events, and allows intervention, that is, the intentional causation or prevention of events. Previous research has shown that in certain contexts, children can learn cause–effect relations merely through observation. This is true if an event is caused by a human being acting in a goal-directed manner, and/or if there is physical contact between the causer and the affected entity. If these conditions are not met, children have more difficulties detecting causal structures. However, studies with English-speaking children have yielded evidence that when such atypical causal relations are described with language, children are more likely to understand them. In this project, we are investigating this effect in the Swiss German language. In Swiss German, causality is predominantly expressed by the lexical causative, that is, with verbs with a causal meaning embedded in a transitive frame. We assess whether Swiss toddlers of 2.5 to 3.5 years of age make use of the transitive verb frame as a cue to causality by comparing a causal language group to a  control condition. Preliminary data reveals a trend that the children in the causal language condition perform somewhat better than the children in the control condition.

Low-income yet highly-educated: Trajectory and achievement differences in children from low-income, highly-educated families
Lauren Tighe, UM
(Advisors: Pamela Davis-Kean, Toni Antonucci)

Past research on children’s achievement often contains samples where parental education and family income are fairly related, without considering families experiencing a disconnect between these socioeconomic indicators. High parental education may have a positive influence on children’s achievement even when living in or near poverty. The purpose of the current study is to examine heterogeneity within low-income yet highly-educated families. Using latent class growth analyses, I determine family’s income status trajectories over time, from kindergarten to eighth grade, in order to differentiate between families experiencing chronic or transient poverty. Then, I examine parental characteristics (e.g., gender, race, marital status, working status) to better understand descriptive differences, or similarities, amongst trajectory classes. Lastly, I examine differences in children’s eighth grade reading and math achievement by trajectory class.

Sentimental norms: Children and adults show greater physiological arousal to moral than conventional transgressions
Meltem Yucel, UVA
(Advisors: Amrisha Vaish, Gerald Clore)

Human group living and cooperation rely heavily on norms, or guidelines that direct how one ought to behave (Boyd & Richerson, 2009; Killen & Smetana, 2015). However, not all norms are perceived similarly or treated the same. In particular, people robustly distinguish between moral norms, which are aimed at preserving the rights and welfare of others (e.g., not causing unprovoked harm or not stealing from others), and conventional norms, which serve to ensure coordination and smooth running of social groups or institutions (e.g., wearing uniforms to school or following certain rules when playing a game) (Davidson, Turiel, & Black, 1983; Smetana, 1984; 1985; Weston & Turiel, 1980). Indeed, even children as young as 3 years of age judge moral norm violations as more serious, moral transgressors as more deserving of punishment, and moral rules as more generalizable across contexts (Smetana, Ball, Jambon, & Yoo, 2018).
However, far less is known about the mechanisms driving this differentiation. One proposal argues that the basis for this differentiation is affect: whereas moral norm violations elicit affect, conventional norm violations do not (Nichols, 2002). Though this proposal has found some support in research with adults (Nichols, 2002), prior work has not examined it in young children. The present study tested this idea in development. Specifically, we asked whether children show greater physiological arousal (measured via changes in pupil dilation) during moral transgressions. We also asked whether children attend differently to moral versus conventional transgressions.
In a between-subjects design, 3-year-olds (n = 32), 4-year-olds (n = 34), and undergraduate students (n = 64) watched a video of either a moral norm violation (e.g., destroying another person’s artwork) or a conventional norm violation (e.g., playing a game wrong). Crucially, the final scene of both videos was identical, but based on what had previously transpired, this final scene represented either a conventional or a moral transgression.
Participants of all age groups attended significantly more to the victim of the moral transgression than the bystander in the conventional transgression, even though both scenes were identical and thus the victim/bystander behaved identically in both, F(1, 124) = 6.19, p = .014. With regard to our central question concerning physiological arousal, we measured participants’ pupil dilation during the first 2 seconds of each transgression. As predicted, we found that across all age groups, participants in the moral condition (M = .10, SD = .15)showed greater pupil dilation than participants in the conventional condition (M = .01, SD = .14), F(1,112) = 11.50, p = .001. We also found a main effect of age, such that adults showed less change in pupil size than 3- and 4-year-olds, F(2,112) = 3.36, p = .038. This is the first evidence that differences in affective arousal contribute to the distinction that even young children draw between moral and conventional rules.