Helmholtz Centre Potsdam - GFZ German Research Centre for Geosciences
PIK PublicationsNot a member yet
55339 research outputs found
Sort by
Justification of Violence 1
No full textThis study is the first study in a larger project. It aims to examine the role of justification of violence on victims and perpetrators of intimate partner violence. We mainly examine its association with commitment and forgiveness
Investigating the maintenance of lexically entrained terms across adulthood
No full textIntroduction
During dialogue, adults tend to imitate their interlocutors’ language use. This conversational alignment occurs at multiple different levels of language and seems to function to ensure effective communication and rewarding interactions. In this study we specifically focus on the tendency for a speaker to reuse the same word as that used by a conversational partner (e.g. using brolly after your partner has used brolly; Brennan & Clark, 1996), known as lexical entrainment.
It is well established that lexical entrainment implicates both unmediated processing (e.g. lexical retrieval) and mediated processing (e.g. speakers’ beliefs about their interlocutor). Unmediated theories have highlighted the role of automatic priming in lexical entrainment, whereby exposure to a partner’s lexical label (e.g. brolly) makes its representation more accessible in memory and easier to retrieve and reuse (Pickering & Garrod, 2004). Moreover, mediated theories have suggested that speakers’ beliefs may also influence their tendency to entrain – they may adapt their use of language based on what they believe a conversational partner will understand (audience design mechanisms; Clark, 1996), or in order to express affiliation and establish rapport (social-affective mechanisms; van Baaren et al., 2003).
Previous studies have examined the mechanisms underlying lexical entrainment using group-level comparisons (e.g. Branigan et al., 2011). However, mediated and unmediated processing mechanisms can also vary as a result of individual differences (Tobar-Henriquez et al., 2019); this thus provides the opportunity to understand how and why lexical entrainment might vary based on individual characteristics such as age, prosociality and language ability. However, in order to understand the role of individual differences in lexical entrainment, we must first establish whether lexical entrainment is a trait which is stable within individuals. We have previously established this in adults aged 18-60 (Tobar-Henríquez et al., 2020) and in the following age groups: 18-39, 40-59, and over 60 (Tobar-Henríquez & Wilks, 2021). Furthermore, previous studies have found that lexical entrainment in adulthood increases with age, so that older speakers entrain more than younger speakers (Tobar-Henríquez, 2020; Tobar-Henríquez & Wilks, 2021). This is an important finding because it suggests that lexical entrainment mechanisms change across adulthood, thus opening up new directions to investigate both lexical entrainment mechanisms, in particular, and language processing across the lifespan, in general.
One possible explanation for our finding of increased lexical entrainment with age (Tobar-Henríquez, 2020; Tobar-Henríquez & Wilks, 2021) is that priming mechanisms underlying entrainment are relied upon more heavily as speakers age. Older adults find retrieving words from memory (lexical retrieval) more difficult than younger adults. For example, they experience more tip-of-the-tongue states, take longer to find the right words (e.g. ‘dog’) for target objects (e.g. a dog), and make more naming mistakes than younger adults, using more semantically-related but inaccurate words for targets (e.g. referring to a dog as ‘cat’; e.g. Burke & Laver, 1990; Mitchell, 1989; Nicholas et al., 1985; Thomas et al., 1977). However, their difficulties seem restricted to accessing the phonological, rather than the semantic, representations of words. This is evidenced by the fact that elderly speakers make correct prime-target semantic associations (e.g. associating dog with cat) and successfully use phonological cues to overcome lexical retrieval difficulties (e.g. successfully recalling dog after being told that the word started with the letter d; Barresi et al., 2000; Juncos-Rabadán et al., 2009; Kavé & Mashal, 2012; Nicholas et al., 1985).
Thus, older speakers should find retrieving a name, that they have not recently processed, more difficult than younger speakers. This may increase the likelihood that they will entrain to the specific term used by their partner as the term has been processed more recently and thus is easier to retrieve, resulting in increased entrainment with age. Yet, it is not known how age may affect whether such an entrained term is later re-used with a new partner. Moving forward, we believe this “maintenance of entrained terms” to be an important consideration because in order to communicate inter-generationally an individual needs to continuously update their language to incorporate changes driven by younger speakers. Lexical entrainment could thus represent a means by which to learn new lexical labels and the correct contexts of use (Tobar-Henríquez et al., 2021).
The extent to which entrained terms are maintained may result from an interplay between lexical priming mechanisms and perspective-taking abilities. Perspective-taking decreases with age (Horton & Spieler, 2007) but reports on lexical priming mechanisms and age are somewhat mixed, with most researchers failing to find an age effect (Barry et al., 2006; Jones & Estes, 2012; Mitchell, 1989; Mitchell et al., 1990; Stern et al., 1991). Additionally, the studies often fail to replicate due to limited participant numbers and a focus on reaction times. Therefore, assuming reduced perspective-taking abilities and intact priming mechanisms, we might expect older adults to maintain use of terms used with one partner, with a new partner, at higher levels than younger adults. This is because understanding that a new partner doesn’t necessarily favour the entrained term, and modulation of priming effects driving entrainment, may require perspective taking. That is, lexical priming mechanisms could be activated by default but potentially over-ridden in certain situations in order to enable more flexible communication. Older adults may therefore display maladaptive over-maintenance of entrained terms (i.e. maintenance that might lead to miscommunication; see converging evidence from young children’s referential language use; Garrod & Clark, 1993).
Social cooperativeness increases with healthy aging such that older people tend to be more interested than younger people in engaging in emotionally meaningful activities/increasing rapport during social interaction. They are also more cooperative and emotionally empathetic than younger adults (Beadle et al., 2015; Blanchard-Fields, 2007; Carstensen et al., 1995). It is therefore possible that older adults might entrain more often than younger adults in order to reduce social distance and increase rapport with their conversational partner. This would implicate prosocial accounts of lexical entrainment. In this case, we would expect there to be a positive association between individual prosocial tendencies and lexical entrainment such that adults who are more prosocial entrain more often.
We have previously noted that, in order to understand age differences in lexical entrainment, it was important to establish the stability of lexical entrainment in young, middle-aged and older adults (Tobar-Henríquez & Wilks, 2021). In a similar vein, if we want to understand age differences in the maintenance of lexically entrained terms, we must first establish whether this maintenance is stable in young, middle-aged and older adults. In the current study, we will therefore first confirm that lexical entrainment (aiming to replicate our previous findings), and maintenance of lexical entrainment (an extension to these findings), are stable traits within three groups of participants aged 18-39, 40-59 and over 60. We will then investigate how age might affect lexical entrainment (aiming to replicate our previous findings) and maintenance of lexical entrainment (an extension to these findings). Additionally, we will perform an exploratory investigation into the effect of prosociality on lexical entrainment by examining whether individual differences in prosociality predict lexical entrainment. In addition to our lexical entrainment instrument, we thus use an instrument which is able to consistently distinguish between individuals’ levels of prosociality (see “The present experiment” section)
Strengths Use Meta-Analysis
No full textThe idea of “strengths” is a key construct in positive psychology and the related idea of “strengths use” has more recently been emphasized in this literature, especially in research on positive organizational psychology. Despite its increasing popularity, we lack a coherent quantitative synthesis of research on strengths use at work to inform the (re)development of theory, future research, and the implementation of interventions to encourage strengths use in organizations.
In this meta-analytic study, we explore three overarching research questions:
1. What is the relationship between strengths use and outcomes related to worker well-being (e.g., perceiving one’s work to be meaningful, positive emotions, work engagement, psychological well-being, job satisfaction, etc.) and work performance (e.g., counterproductive behaviors, proactive behaviors, helping behaviors, task performance, etc.)
2. Is there a “Goldilocks zone” of strengths use that implies a non-linear strengths underuse/overuse (inverse U-shaped) curve?
3. What methodological and substantive variables moderate the relationship between strengths use and job performance and well-being
Preregistration of stimulus annotations
No full textPreregistration of manual annotations for our stimuli. This component of our OSF project contains the annotations for all our stimulus files created using ELAN (version 5.7-FX) in the folder "sentences". The folder contains ELAN files (*.eaf and *.pfsx) as well as the video files (*.mp4) of our stimulus sentences
Examining the neural correlates of preference learning and their relationship to social, adaptive, and academic performance in autistic and non-autistic adolescents
No full textAutism spectrum disorder (ASD) is a neurodevelopmental condition defined by core deficits in the areas of social interaction and restricted interests (American Psychiatric Association, 2013). Prior work has documented large-scale differences in brain regions and functional networks linked to social cognition, cognitive flexibility, and learning in autism, including the frontoparietal networks, temporoparietal junction (TPJ), medial prefrontal cortex (MPFC), anterior cingulate cortex (ACC), and insula (Dajani & Uddin, 2015; Gotts et al., 2012; Sami et al., 2023; Y.-J. Yang et al., 2015). With increasing social demands during adolescence (Bechara et al., 2019; Chein et al., 2011; Larson & Richards, 1991; O’Brien et al., 2011), alongside ongoing neurodevelopmental changes (S. J. Blakemore & Robbins, 2012; S.-J. Blakemore, 2008; Ciranka & van den Bos, 2021; Steinberg & Morris, 2000; Towner et al., 2023), autistic adolescents encounter distinct challenges that may shape social and vocational outcomes (Biggs & Carter, 2016; Wallace et al., 2017). Yet, the significant heterogeneity in autistic phenotypes has hindered the development of mechanistic evaluation procedures and individualized interventions that meet the unique needs of autistic individuals (Jones & Klin, 2009; Masi et al., 2017; Ring et al., 2008). Moreover, variability in presentation appears to scale with diverging outcomes rooted in the transition from adolescence to adulthood, underscoring the need to probe the mechanistic roots of autistic heterogeneity (Biggs & Carter, 2016; Kraper et al., 2017).
Paradigms probing social knowledge and learning offer neurocognitive insights into heterogeneity across core autism feature domains.Performance in learning tasks more generally have been shown to scale with variability in repetitive and restricted behaviors (Sapey-Triomphe et al., 2022; South et al., 2012; Zalla et al., 2009) and performance social learning paradigms specifically predict social abilities in ASD (DeMayo et al., 2019; McNaughton et al., 2023; Rosenblau et al., 2021). As such, social learning offers promise for clinical insights into clinical profiles. This study will leverage a previously implemented peer preference learning paradigm (Frolichs et al., 2022; Rosenblau et al., 2018, 2021, 2023) which has been found to scale with social adjustment and behavioral rigidity in autistic adolescents (Rosenblau et al., 2021).
During the task, participants rate how much they think an adolescent likes a series of food and activity items while receiving trial-wise feedback. A key metric of participants’ learning is the prediction error (PE) – in this case, the absolute difference between the presented adolescents’ actual rating and the participants’ estimate. Prior work using this paradigm found that non-autistic participants relied on prior knowledge about average peer preferences as a reference point. Over time, though, participants updated their representations of the adolescent by integrating PEs scaled by item-level similarity resulting in a reduction in PEs overall. Autistic adolescents, on the other hand, solely applied their self-preferences as a reference point for inferences (Rosenblau et al., 2021).
A subsequent study using this general paradigm leveraged significantly larger samples in order to thoroughly characterize the self-preferences of autistic and non-autistic adolescents and examine whether autistic adolescents were instead applying knowledge about other autistic peers. We found that autistic adolescents did reduce PEs over time. Autistic adolescents’ initial expectations – a model derived parameter akin to reference points – scaled with mean autistic preferences. Both groups also scaled PEs by fine-grained item-similarity structures derived from self-preference relationships of their respective groups (Cahalan et al., 2025b).
In the original study (Rosenblau et al., 2021), non-autistic adolescents encoded model-free PEs in the medial prefrontal cortex (MPFC) and model-based PEs in the putamen. In contrast, autistic adolescents did not show PE-related neural activity and instead encoded their self preferences in the angular gyrus (Rosenblau et al., 2021). The MPFC has also been generally implicated in mental state attribution (Adolphs, 2009; Denny et al., 2012; C. Frith & Frith, 2006, 2012; Rosenblau et al., 2023) and atypical recruitment of this region in autism is thought to contribute to mentalizing difficulties in ASD (U. Frith, 1994; Rosenblau et al., 2023; Watanabe et al., 2012; Y-J Yang et al., 2015). It is therefore a primary region of interest for social learning and differences in the neural correlates of social learning in autism.
In light of more recent behavioral findings in larger samples wherein autistic adolescents were incorporating PEs and accounting for prior social knowledge (Cahalan et al., 2025b), the neural correlates of preference learning in non-autistic and autistic adolescents should be re-examined at both the whole brain level and in a pre-specified MPFC region-of-interest (See Aim 1).
In the replication study, mean preferences of both non-autistic and autistic samples were highly correlated with each group’s respective self-preferences introducing ambiguity around which reference point, or initial expectation, more strongly influenced participants’ ratings. From a neural perspective, participants may access distinct neural substrates when inferring based on their self- versus respective average group preferences (i.e. the mean preference of the larger reference population, called mean preferences hereafter). Self-referential thinking has been shown to engage the ventromedial prefrontal cortex (vmPFC, D’Argembeau, 2013; Denny et al., 2012; Hiser & Koenigs, 2018; Konu et al., 2020). Thus, vmPFC could be implicated when participants use their self-preferences.
We also aim to investigate whether brain regions encode coarse- and fine-grained social knowledge using multivariate representational similarity analysis (RSA, Kriegeskorte et al., 2008). This technique allows us to compare representations from behavioral data with the representations from fMRI data. RSA has been widely applied across various subfields of cognitive neuroscience and in the social domain, specifically, it has revealed neural activity patterns underlying the representation of relevant task dimensions (Peer et al., 2021; Popal et al., 2019; Riberto et al., 2022; Stolier et al., 2020). So far, however, studies using RSA have only focused on low-dimensional representations of social information such as personality (Hassabis et al., 2014), or three to five dimensional models for organizing mental states (Tamir et al., 2016) or personality traits (Thornton & Mitchell, 2017, 2018). In our previous social learning study, the Big-5 social learning model was the coarsest model in our task set, and we found that a fine-grained item-level model was better at capturing how participants learned about the preferences of real people (Cahalan et al., 2025b, Frolichs et al., 2022). We therefore hypothesized that along with coarse-grained knowledge representations, people also represent more nuanced semantic structure, such as finer grained item categories and even item-level similarity in cortical activity patterns, especially in the MPFC.
The hippocampus is thought to represent conceptual knowledge in a map-like fashion akin to physical space representations for navigation (Mack et al., 2016; Schafer & Schiller, 2018a). The region has recurrent connections with PE-encoding regions like MPFC (Koster et al., 2018) and the putamen (Starr et al., 2011) suggesting potential for mutual involvement in mapping social representations such as mean preferences (Banker et al., 2021; Kennedy & Courchesne, 2008; Rosenblau et al., 2023; Schafer & Schiller, 2018b). Examining how mean and self-preferences are neurally encoded as participants make their ratings, then, could provide insight into how these reference points support autistic versus non-autistic adolescents’ inferences (See Aim 2).
Furthermore, large cohort studies examining the relationship between childhood poverty and brain maturation have implicated hippocampus development as predictive of academic achievement (Hair et al., 2015). When learning a novel lexicon, relative hippocampal activity has also been shown to predict learning outcomes (Breitenstein et al., 2005). No studies to date have specifically explored the neurocognitive relationship between social learning and academic achievement. Therefore, this study will test whether PE-signals during the social learning task in the hippocampus could serve as neurocognitive predictors of academic/intellectual abilities (H3.4).
Crucially, findings from Cahalan et al., 2025b demonstrated that variability in performance metrics and model parameters within this social learning framework mapped on to variability in autistic profiles. Autistic adolescents reporting elevated behavioral rigidity over and above composite autistic traits exhibited an attenuated reduction in PEs–or less learning over time–and inferred their peers would rate items lower at the start of the task. In a weaker association, autistic adolescents also reporting more autistic traits overall showed lower learning rates. Growing evidence suggests that factors like age (Jones et al., 2014; Rosenblau et al., 2018; Towner et al., 2023), autistic trait load (Haffey et al., 2025), and social abilities (McNaughton et al., 2023; Rosenblau et al., 2020, 2021) may account for variability in task-related neural activity in paradigms examining social cognition which may not be readily apparent from group-level comparisons. This effect may be most apparent in regions implicated in social PE encoding and mental state attribution such as the MPFC (Ibrahim et al., 2021; Rosenblau et al., 2020).
Given the role of social learning for predicting autism symptomatology and its relevance for learning in real world social and academic settings (Rodriguez Buritica et al., 2019; Rosenblau et al., 2020, 2023; Towner et al., 2023), this study aims to examine whether variability in neural activity associated with preference learning can be traced to broader individual differences in symptom profiles and demographic factors and predict academic and social outcomes
To Protest (Anger) or Not to Protest (Fear): A New Emotional Dissident Perspective
No full textCauses of protest continuatio
Sites of Meaning Making: Study 3a
No full textIn this study, I will manipulate whether remembering a place you are attached to (versus a non-attached place) contributes to higher meaning in life. I plan to explore the idea of attachment by asking participants to remember familiar places they are attached to or neutral about. Afterwords, I will ask them to answer questions about their personality and the self
Remote monitoring for long-term physical health conditions: Protocol for an evidence and gap map
No full textRemote monitoring allows health care professionals to assess and manage the health of a patient, without the need for the patient to be seen face-to-face. This offers opportunities for innovation in the delivery of health care as well as having the potential to reduce health care costs. Access to current evidence on the most effective forms of remote monitoring, and factors affecting their acceptability and implementation, will support health care providers in delivering these interventions. We aim to identify, classify, and appraise recent systematic reviews of the effectiveness of remote monitoring, and its acceptability and implementation, for people living with long term physical health conditions, to produce an evidence and gap map
Politicians' evaluation of public opinion - module 3 - 2022 data collection
No full textIn democracies, policies are expected to be responsive to public opinion. Extant research showed that responsiveness is selective. It varies across issues, time and countries. Yet, how come policies vary in their responsiveness has not received a satisfying answer. This project examines the puzzle of why policy responsiveness varies. Its core argument holds that politicians evaluate public opinion and let their actions—in line with public opinion or going against it—depend on their appraisal. When public opinion is evaluated negatively, it has no effect on what politicians do; that it is evaluated positively increases the chance that politicians act congruently. The project examines three matters: (1) which criteria politicians use to appraise public opinion; (2) how, depending on the opinion content of the message, the channel through which the opinion is conveyed and the group from which it comes, concrete public opinion signals are evaluated; and, (3) which effect these evaluations have on politicians’ political action. The central expectation is that public opinion is evaluated by politicians based on a consistent and common scoreboard. For instance, opinion signals are rated based on their representativity and underlying public opinion is evaluated on its quality and its intensity. The project tackles these matters drawing on a comparative study in fourteen countries (Australia, Belgium (both Flanders and Francophone Belgium), Canada, Czech Republic, Denmark, Germany, Israel, Luxembourg, Netherlands, Norway, Portugal, Switzerland, and Sweden).
This registration relates to the first wave of data collection of *anonymous project*, to be carried out in 2022 and starting on March 21st, 2022. In this round, seating national and regional politicians in the fourteen countries will be surveyed and interviewed with regard to how they evaluate public opinion and the behavioral consequences thereof. There will be a second round of data gathering among politicians sometime in 2024 or 2025. A separate registration will be done for that second round of data.
Note that the 2022 questionnaire contains a number of other questions, measures and experiments that will be registered separately. This registration only relates to the questions with regard to politicians' evaluation of public opinion
Epistemic Cooperation: Experiment 4
No full textIn this study, participants play a 1-player trivia game with the same overarching design as in previous studies. Participants are sorted into one of two conditions – an “Easy” trivia condition and a “Hard” trivia condition. In the Easy condition, participants answer very basic trivia questions that are designed such that all participants will know the answer prior to seeing their partner’s response (e.g. how many fingers does a typical human have on one hand?) In the Hard trivia condition, questions are designed to tap into liberals’ and conservatives’ distinctive cultural knowledge bases, generating opportunities for partisans to “rescue” their out-party partners.
The goal of this analysis is to ask whether effects on out-party attitudes depend on actually being helped by an outgroup member’s knowledge