14 research outputs found
Recommended from our members
The Self, Emotions, and Self-Injury: Links Between Self-Oriented Cognition and Emotion Processes in Non-Suicidal Self-Injury
No full textNon-suicidal self-injury (NSSI)—the intentional destruction of one’s own body tissue without suicidal intent (e.g., self-cutting, self-hitting; Hooley et al., 2020)—is widely prevalent, occurs across different psychiatric disorders, and elevates risk for suicide. Decades of research indicate that self-oriented cognition (e.g., attentional focus on the self; our beliefs about ourselves) and emotion processes (e.g., how we experience and regulate emotions) are key risk and maintenance factors for NSSI. For example, a fluctuating and predominantly negative self-concept is strongly associated with NSSI onset and maintenance, as are difficulties understanding, tolerating, and regulating negative emotions. Although self-oriented cognition and emotion processes are dually implicated in NSSI, these facets are typically studied separately in NSSI research. We therefore know remarkably little about how self-concept impairments are linked to emotion impairments within people struggling with NSSI. This gap in knowledge is concerning given that our beliefs about ourselves are associated with our emotions, and risk factors for psychopathology tend to interact.
Across three studies, this dissertation examines relationships between self-oriented cognition and emotion processes within people who engage in NSSI. In study 1 (N = 134; ages 18+), we developed a lab-based behavioral task to examine how adults who engage in NSSI (n=40) regulate negative affect when experiencing negative self-focus, compared with adults who engage in indirect forms of self-injury (e.g., difficulties with substance use, disordered eating; n=46) and adults with no self-injurious behaviors (n=48). People who engage in NSSI and indirect forms of self-injury alike were more likely than those without self-injury to select behavioral strategies (e.g., a word activity, physical pain) relative to cognitive strategies (e.g., reappraisal, rumination) to regulate their emotions. Findings suggest that people who engage in NSSI are more likely to choose behavioral regulation strategies when experiencing negative self-focus, even when cognitive strategies are available and made salient.
Study 2 (N = 286) examines how emerging adults (ages 18-25) who engage in NSSI regulate negative emotions that require self-evaluation (shame and guilt) versus negative emotions that do not (sadness and anger). Participants with and without NSSI (n = 150 and n=136, respectively) completed an online behavioral task that induced specific negative emotions through autobiographical memories. Following autobiographical memory recall, participants selected one of three cognitive emotion regulation strategies: rumination (thinking further about the memory), distraction (thinking about something unrelated to the memory) and counteract (recalling a positive autobiographical memory that elicits pride). People who engage in NSSI were more likely than those without NSSI to ruminate when experiencing sadness or anger, and to distract or counteract when experiencing guilt. These findings suggest that discrete negative emotions may be linked to NSSI via different emotion regulation pathways.
In Study 3 (N = 187) we explored associations among multiple components of self-concept and emotion that have each been associated with NSSI. At two time points across a three-month span, emerging adults (ages 18-25) with and without a history of NSSI (wave 1: n=87 and n = 100, respectively; wave 2: n = 44 and n = 42) completed questionnaires assessing dimensions of self-concept (e.g., self-criticism, self-alienation) and emotion (e.g., difficulties identifying emotions, emotion regulation tendencies). Regularized network analyses were used to compare relationships within and between groups across time. Results indicated that self-alienation and difficulties identifying emotions were among the most strongly connected variables to other variables in the network for people with and without a history of NSSI alike, both at wave 1 and wave 2. Accordingly, feeling disconnected from oneself and being unable to identify one’s emotions may lead to additional self- and emotion-related difficulties for people with and without NSSI.
Taken together, these studies indicate that examining self-oriented cognition and emotion processes concurrently could deepen our understanding of NSSI. Results further highlight the importance of considering self and emotion processes jointly in NSSI treatment
Emotion Regulation Strategy Choices Following Aversive Self-Awareness in People Who Engage in Nonsuicidal Self-Injury or Indirect Self-Injury
No full textEmotion regulation difficulties are implicated prominently in self-injury. Yet it is unclear how people who engage in different forms of self-injury attempt to regulate negative affect when multiple strategies are available to them. This laboratory-based study examined emotion regulation strategy choices in individuals who engage in non-suicidal self-injury (n=40), indirect forms of self-injury (disordered eating and substance abuse; n=46), and controls (n=48). Following a self-relevant stressor (negative autobiographical memory recall), participants selected one of six regulation strategies based on what they believed would most effectively alter their affect. Strategies spanned behavioral (physical pain, a snack, word activity) and non-behavioral (rumination, reappraisal, doing nothing) domains. Compared to controls, individuals who engage in NSSI and indirect self-injury were more likely to select behavioral strategies. In addition, those with NSSI and indirect self-injury were more likely than controls to choose physical pain and less likely to ruminate. Findings indicate that people with direct and indirect forms of self-injury alike are more likely to take action than to engage in further thought when experiencing aversive self-awareness, even when cognitive strategies are made salient. Results illuminate intervention targets for these clinical populations
Recommended from our members
The Relationship Between Perceived Criticism and Trait Victimhood
No full textThose who perceive that people are more critical of them also tend to feel more victimized in their relationships. This relationship was measured with the Perceived Criticism scale which reliably predicts relapse and recurrence of symptoms in mental health disorders, and the Tendency for Interpersonal Victimhood (TIV) scale. TIV is a newly developed personality construct that measures an individual’s ongoing sense of victimhood that permeates many different types of interactions. The participants of the study were 242 individuals who were between 18 and 30 years old and they were recruited by the Clinical Research Lab from the Harvard university pool and online platform Reddit. Pearson’s correlation showed a significant relationship between perceived criticism and the tendency to feel victimhood. Additionally, the dimensions of the TIV (the need for recognition, moral elitism, lack of empathy, and rumination) were also correlated with the perceptions of criticism. Those who evaluated their close ones as highly critical had higher rates of rumination, felt lack of empathy, needed more recognition, and felt morally superior. The sample for this study consisted mostly of students; therefore, this relationship between perceived criticism and victimhood can be explored further with participants from clinical settings using different demographics to obtain more generalizable results and compare PC average scores across different cultures
Understanding the self in nonsuicidal self-injury: A conceptual review and framework for future research
No full textNonsuicidal self-injury (NSSI) is a widespread and dangerous behavior. Despite increasing understanding of the risk factors for NSSI, this behavior remains highly prevalent, highlighting the need for an extension of current research and more precise treatment targets. Specifically, research examining self-perception in NSSI provides a fruitful foundation for future work. Mounting studies indicate that self-concept disturbances are implicated in NSSI. Yet it remains unclear how different components of self-concept—such as self-criticism and identity confusion—are associated for people with NSSI. Furthermore, research in this domain uses distinct definitions and measures of self-concept disturbance, rendering it difficult to integrate findings across studies. This conceptual review provides the first summary to date synthesizing research on self-concept (content and structure) in NSSI, highlights research questions to address, and outlines suggestions for future work. Recommendations for NSSI research examining self-concept include: (a) increasing the consistency of terms used; (b) examining relationships between self-concept content and structure; (c) exploring the extent to which measures of self-concept and identity tap into the same phenomena; (d) assessing self-concept across different levels of analysis; and (e) identifying treatment targets for distinct self-related disturbances (e.g., heightened self-criticism versus an inconsistent sense of self). Potential intervention targets are discussed
The perceptual primacy of feeling: Affectless machine vision models explain a majority of variance in human visually-evoked affect
No full textLooking at the world often involves not just seeing things, but feeling things. Modern feedforward machine vision systems that learn to perceive the world in the absence of active physiology, deliberative thought, or any form of feedback that resembles human affective experience offer tools to demystify relationship between seeing and feeling, and to assess how much of visually-evoked affective experiences may be a straightforward function of representation learning over natural image statistics. In this work, we deploy a diverse sample of 180 state-of-the-art deep neural network models trained only on canonical computer vision tasks to predict human ratings of arousal, valence, and beauty for images from multiple categories (objects, faces, scenescapes, art) across two datasets. Importantly, we use the features of these models without additional learning, linearly decoding human affective responses from network activity in much the same way neuroscientists decode information from neural recordings. Aggregate analysis across our survey demonstrates that predictions from purely perceptual models explain a majority of the explainable variance in average ratings of arousal, valence, and beauty alike. Finer-grained analysis within our survey (e.g. comparisons between shallower and deeper layers, or between randomly-initialized, category-supervised, and self-supervised models) point to rich, pre-conceptual abstraction (learned from diversity of visual experience) as a key driver of these predictions. Taken together, these results provide further computational evidence for an information-processing account of visually-evoked affect linked directly to efficient representation learning over natural image statistics, and hint at a computational locus of affective and aesthetic valuation immediately proximate to perception
The perceptual primacy of feeling: Affectless machine vision models explain a majority of variance in human visually-evoked affect
No full textLooking at the world often involves not just seeing things, but feeling things. Modern feedforward machine vision systems that learn to perceive the world in the absence of active physiology, deliberative thought, or any form of feedback that resembles human affective experience offer tools to demystify relationship between seeing and feeling, and to assess how much of visually-evoked affective experiences may be a straightforward function of representation learning over natural image statistics. In this work, we deploy a diverse sample of 180 state-of-the-art deep neural network models trained only on canonical computer vision tasks to predict human ratings of arousal, valence, and beauty for images from multiple categories (objects, faces, scenescapes, art) across two datasets. Importantly, we use the features of these models without additional learning, linearly decoding human affective responses from network activity in much the same way neuroscientists decode information from neural recordings. Aggregate analysis across our survey demonstrates that predictions from purely perceptual models explain a majority of the explainable variance in average ratings of arousal, valence, and beauty alike. Finer-grained analysis within our survey (e.g. comparisons between shallower and deeper layers, or between randomly-initialized, category-supervised, and self-supervised models) point to rich, pre-conceptual abstraction (learned from diversity of visual experience) as a key driver of these predictions. Taken together, these results provide further computational evidence for an information-processing account of visually-evoked affect linked directly to efficient representation learning over natural image statistics, and hint at a computational locus of affective and aesthetic valuation immediately proximate to perception
Comparing self‐harming intentions underlying eating disordered behaviors and NSSI: Evidence that distinctions are less clear than assumed
No full textAttenuated beta-adrenergic response to stress and increased anticipation and perception of social threat in women high on perceived criticism
No full textA large amount of literature has demonstrated that Perceived Criticism (PC)-that is, how critical a person believes a given relative is of him or her-is associated with negative clinical outcomes in a broad range of psychiatric disorders (e.g., relapse or recurrence of symptoms). A possible mechanism behind the predictive value of PC might be its association with the stress regulation process. This is the first study to investigate differences in the psychophysiological response to a social stress task in young women (mean age = 21.66, SD = 4.33) with high (n = 40) and low (n = 39) PC. The physiological response was investigated by measuring two markers of sympathetic activity mediated by acetylcholine (skin conductance levels; SCL) and adrenaline (preejection period; PEP) levels, respectively, and one marker of the vagally-mediated parasympathetic system (heart rate variability; HRV). Moreover, we investigated the anticipation and perception of social threat, in the form of criticism, during the stressor. No differences in HRV and SCL were observed. However, individuals high in PC mobilized fewer cardiovascular resources to deal with the stressor, reflected in an attenuated beta-adrenergic response (i.e., lower PEP response). Women high in PC also expected and perceived more criticism during the stress task. Together, our results indicate that women high in PC make heightened social threat anticipation and interpretations, and they tend to engage in less active coping when exposed to socially evaluated stressful events. Our findings indicate that PC is associated with underlying stress-related psychobiological vulnerabilities that may contribute to its association with negative clinical outcomes
Using Multimodal Deep Neural Networks to Disentangle Language from Visual Aesthetics
No full textWhen we experience a visual stimulus as beautiful, how much of that experience derives from perceptual computations we cannot describe versus conceptual knowledge we can readily translate into natural language? Disentangling perception from language in visually-evoked affective and aesthetic experiences through behavioral paradigms or neuroimaging is often empirically intractable. Here, we circumnavigate this challenge by using linear decoding over the learned representations of unimodal vision, unimodal language, and multimodal (language-aligned) deep neural network (DNN) models to predict human beauty ratings of naturalistic images. We show that unimodal vision models (e.g. SimCLR) account for the vast majority of explainable variance in these ratings. Language-aligned vision models (e.g. SLIP) yield small gains relative to unimodal vision. Unimodal language models (e.g. GPT2) conditioned on visual embeddings to generate captions (via CLIPCap) yield no further gains. Caption embeddings alone yield less accurate predictions than image and caption embeddings combined (concatenated). Taken together, these results suggest that whatever words we may eventually find to describe our experience of beauty, the ineffable computations of feedforward perception may provide sufficient foundation for that experience
