Ying Li
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Previous Organizations / Institutes
The organizations / institutes from the scholar's other articles in the past ten years
Research Subjects
Ying Li has a total of 107 co-authors. Here are the top co-authors by number of shared publications:
Thomas Hills
9 shared publications- Massimo Stella
CogNosco Lab, Department of Psychology and Cognitive Science, University of Trento, 38068 Rovereto, Italy
6 shared publications Ralph Hertwig
5 shared publications- Thomas Trenholm Hills
Department of Psychology, University of Warwick
4 shared publications Andreia Sofia Teixeira
3 shared publications- Cynthia S. Q. Siew
Department of Psychology, National University of Singapore, Singapore
3 shared publications
Latest Publications
- Open Access
- 20 March 2026
- Published by National Academy of Sciences in Proceedings of the National Academy of Sciences
- Vol. 123 (12)
The private solution trap in collective action problems across 34 nations
Collective action problems emerge when individual incentives and group interests are misaligned, as in the case of climate change. Individuals involved in these problems are generally considered to have two options: contribute toward public solutions such as global warming mitigation or free ride. However, many collective action problems today involve a third option of investing in a “private solution” such as local adaptation. The availability of this third option can lead to a private solution trap whereby private solutions are adopted, collectively optimal public solutions are not provided, and existing inequalities are exacerbated. We investigated the private solution trap with a collective action game featuring private and public solutions, wealth inequality determined by luck or merit, and participants from 34 countries. We found that the joint existence of private solutions and wealth inequality had a consistent effect across countries: Participants given a higher endowment adopted private solutions almost twice as often as those given a lower endowment, regardless of whether it was determined by luck or merit, and contributed proportionally less toward public solutions. Wealth inequality increased in every country and those given lower endowments were often left unprotected as public solutions were not provided. Across countries, cultural values of hierarchy and harmony were associated with preferences for private and public solutions, respectively. We also identified two universal pathways toward public solution provision: early contributions and conditional cooperation. Our findings highlight the ubiquity of the private solution trap, its cultural underpinnings, and its potential consequences for global collective action problems.
- 20 March 2026
- Published by National Academy of Sciences in Proceedings of the National Academy of Sciences
- Vol. 123 (12)
Open AccessThe private solution trap in collective action problems across 34 nations
Collective action problems emerge when individual incentives and group interests are misaligned, as in the case of climate change. Individuals involved in these problems are generally considered to have two options: contribute toward public solutions such as global warming mitigation or free ride. However, many collective action problems today involve a third option of investing in a “private solution” such as local adaptation. The availability of this third option can lead to a private solution trap whereby private solutions are adopted, collectively optimal public solutions are not provided, and existing inequalities are exacerbated. We investigated the private solution trap with a collective action game featuring private and public solutions, wealth inequality determined by luck or merit, and participants from 34 countries. We found that the joint existence of private solutions and wealth inequality had a consistent effect across countries: Participants given a higher endowment adopted private solutions almost twice as often as those given a lower endowment, regardless of whether it was determined by luck or merit, and contributed proportionally less toward public solutions. Wealth inequality increased in every country and those given lower endowments were often left unprotected as public solutions were not provided. Across countries, cultural values of hierarchy and harmony were associated with preferences for private and public solutions, respectively. We also identified two universal pathways toward public solution provision: early contributions and conditional cooperation. Our findings highlight the ubiquity of the private solution trap, its cultural underpinnings, and its potential consequences for global collective action problems.
- Open Access
- 8 November 2025
- Published by Center for Open Science in PsyArXiv
Older yet Faster: Representational Development Explains Accelerated Semantic Processing in Older Adults
Slower responses in older age are often taken as evidence of cognitive decline, yet they may also stem from representational changes shaped by lifelong learning. We advance a developmental perspective on aging, proposing that behavioral change across the lifespan arises from representational systems continuously updated through learning from the environment. This view predicts that aging may bring both benefits and costs: Performance improves when task demands are congruent with accumulated knowledge but declines when they conflict. We provide convergent evidence for this account across three studies, each comprising the following components. First, in tasks requiring participants to access word meanings (i.e., when task demands were congruent with representation), we found that older adults (70–80 years old) exhibited faster processing speed than younger adults (18–25 years old)—a finding that cannot be explained by dominant decline-based accounts. Second, we traced this processing advantage to systematic changes in older adults’ conceptual representations, as inferred from semantic rating tasks. Third, using learning simulation models, we demonstrated how such representational changes could emerge naturally from cumulative language experience. Replication across different samples, task requirements, and word stimuli strengthens the robustness of these findings. Finally, we identified a boundary condition: when learned representations contained information incongruent with task demands, older adults’ greater learning experience not only attenuates their representational changes but also reduces processing-speed advantage. Altogether, our research challenges the explanatory framework of cognitive aging that emphasizes decline and compensation, advancing a more balanced view grounded in representational development and lifelong learning.
- 8 November 2025
- Published by Center for Open Science in PsyArXiv
Open AccessOlder yet Faster: Representational Development Explains Accelerated Semantic Processing in Older Adults
Slower responses in older age are often taken as evidence of cognitive decline, yet they may also stem from representational changes shaped by lifelong learning. We advance a developmental perspective on aging, proposing that behavioral change across the lifespan arises from representational systems continuously updated through learning from the environment. This view predicts that aging may bring both benefits and costs: Performance improves when task demands are congruent with accumulated knowledge but declines when they conflict. We provide convergent evidence for this account across three studies, each comprising the following components. First, in tasks requiring participants to access word meanings (i.e., when task demands were congruent with representation), we found that older adults (70–80 years old) exhibited faster processing speed than younger adults (18–25 years old)—a finding that cannot be explained by dominant decline-based accounts. Second, we traced this processing advantage to systematic changes in older adults’ conceptual representations, as inferred from semantic rating tasks. Third, using learning simulation models, we demonstrated how such representational changes could emerge naturally from cumulative language experience. Replication across different samples, task requirements, and word stimuli strengthens the robustness of these findings. Finally, we identified a boundary condition: when learned representations contained information incongruent with task demands, older adults’ greater learning experience not only attenuates their representational changes but also reduces processing-speed advantage. Altogether, our research challenges the explanatory framework of cognitive aging that emphasizes decline and compensation, advancing a more balanced view grounded in representational development and lifelong learning.
- Open Access
- 21 September 2025
- Published by Wiley in Psych Journal
- Vol. 14 (6) , 940-951
Enhancing the Cohesion and Influence of Minority Opinions Through Clustering: A Social Network Experiment
Minority opinions can be of crucial importance to the diversity, productivity, and harmony of a group, but are often left unattended and unheard. Previous methods that tried to enhance minority influence are usually overly forceful and low on ecological validity. To overcome these pitfalls, we proposed a new intervention method called minority clustering and examined its effects with a social network experiment ( N = 456). Minority clustering was implemented by increasing the network connections among participants with initial opinions that deviated from the mainstream opinion and forming an opinion cluster among these minority members. Our results show that minority clustering significantly slowed down the rate at which minority members shifted toward majority opinions, thereby sustaining minority cohesion, and moved majority members closer to minority opinions, thus enhancing minority influence. An additional filter bubble intervention, through which all members of a network were exposed to neighbors with similar opinions to their own, further strengthened minority cohesion but weakened minority influence. Minority clustering is an unobtrusive intervention that does not need overt cooperations of network members and can be implemented easily in social media platforms. The working mechanisms of minority clustering and its effects on group opinion formation are further discussed.
- 21 September 2025
- Published by Wiley in Psych Journal
- Vol. 14 (6) , 940-951
Open AccessEnhancing the Cohesion and Influence of Minority Opinions Through Clustering: A Social Network Experiment
Minority opinions can be of crucial importance to the diversity, productivity, and harmony of a group, but are often left unattended and unheard. Previous methods that tried to enhance minority influence are usually overly forceful and low on ecological validity. To overcome these pitfalls, we proposed a new intervention method called minority clustering and examined its effects with a social network experiment ( N = 456). Minority clustering was implemented by increasing the network connections among participants with initial opinions that deviated from the mainstream opinion and forming an opinion cluster among these minority members. Our results show that minority clustering significantly slowed down the rate at which minority members shifted toward majority opinions, thereby sustaining minority cohesion, and moved majority members closer to minority opinions, thus enhancing minority influence. An additional filter bubble intervention, through which all members of a network were exposed to neighbors with similar opinions to their own, further strengthened minority cohesion but weakened minority influence. Minority clustering is an unobtrusive intervention that does not need overt cooperations of network members and can be implemented easily in social media platforms. The working mechanisms of minority clustering and its effects on group opinion formation are further discussed.
- Open Access
- 27 June 2025
- Published by MDPI AG in Big Data and Cognitive Computing
- Vol. 9 (7) , 171
Cognitive Networks and Text Analysis Identify Anxiety as a Key Dimension of Distress in Genuine Suicide Notes
Understanding the mindset of people who die by suicide remains a key research challenge. We map conceptual and emotional word–word co-occurrences in 139 genuine suicide notes and in reference word lists, an Emotional Recall Task, from 200 individuals grouped by high/low depression, anxiety, and stress levels on DASS-21. Positive words cover most of the suicide notes’ vocabulary; however, co-occurrences in suicide notes overlap mostly with those produced by individuals with low anxiety (Jaccard index of 0.42 for valence and 0.38 for arousal). We introduce a “words not said” method: It removes every word that corpus A shares with a comparison corpus B and then checks the emotions of “residual” words in . With no leftover emotions, A and B are similar in expressing the same emotions. Simulations indicate this method can classify high/low levels of depression, anxiety and stress with 80% accuracy in a balanced task. After subtracting suicide note words, only the high-anxiety corpus displays no significant residual emotions. Our findings thus pin anxiety as a key latent feature of suicidal psychology and offer an interpretable language-based marker for suicide risk detection.- 27 June 2025
- Published by MDPI AG in Big Data and Cognitive Computing
- Vol. 9 (7) , 171
Open AccessCognitive Networks and Text Analysis Identify Anxiety as a Key Dimension of Distress in Genuine Suicide Notes
Understanding the mindset of people who die by suicide remains a key research challenge. We map conceptual and emotional word–word co-occurrences in 139 genuine suicide notes and in reference word lists, an Emotional Recall Task, from 200 individuals grouped by high/low depression, anxiety, and stress levels on DASS-21. Positive words cover most of the suicide notes’ vocabulary; however, co-occurrences in suicide notes overlap mostly with those produced by individuals with low anxiety (Jaccard index of 0.42 for valence and 0.38 for arousal). We introduce a “words not said” method: It removes every word that corpus A shares with a comparison corpus B and then checks the emotions of “residual” words in . With no leftover emotions, A and B are similar in expressing the same emotions. Simulations indicate this method can classify high/low levels of depression, anxiety and stress with 80% accuracy in a balanced task. After subtracting suicide note words, only the high-anxiety corpus displays no significant residual emotions. Our findings thus pin anxiety as a key latent feature of suicidal psychology and offer an interpretable language-based marker for suicide risk detection. - Open Access
- 28 December 2023
- Published by Center for Open Science in PsyArXiv
Cognitive networks and text analysis identify anxiety as a key dimension of psychological distress in suicide notes
Understanding the mindset of people who end their lives by suicide remains a key research challenge. To this aim, we reconstruct conceptual associations and emotional perceptions of 139 authors of genuine suicide notes, i.e., notes written moments before ending one's own life. Our methods are grounded in cognitive network science, text analysis and psychometrics. We introduce a quantitative framework measuring to what extent authors of suicide notes tend to associate emotional words in ways similar to how 200 individuals with high/low levels of depression, anxiety, and stress (on a DASS-21 scale) recall concepts together. In Study 1, we use text analysis to build one co-occurrence network out of suicide notes' texts and six fluency co-occurrence networks out of fluency data from individuals high/low on anxiety/stress/depression levels. We find that emotional word associations in suicide notes reflect mostly conceptual associations produced by low-anxiety individuals. In Study 2, we perform residual emotional profiling, i.e. measuring the emotional intensity of words remaining after removing from fluency networks those words mentioned and connected in the suicide notes' network. We find that only the high-anxiety fluency network displays negligible emotional residuals, indicating the highest overlap - in terms of emotions - between suicide notes' content and the concepts recalled by individuals with high anxiety levels. We discuss how these findings relate to existing psychological literature, along with their potential implications for understanding and quantifying suicidal behavior.
- 28 December 2023
- Published by Center for Open Science in PsyArXiv
Open AccessCognitive networks and text analysis identify anxiety as a key dimension of psychological distress in suicide notes
Understanding the mindset of people who end their lives by suicide remains a key research challenge. To this aim, we reconstruct conceptual associations and emotional perceptions of 139 authors of genuine suicide notes, i.e., notes written moments before ending one's own life. Our methods are grounded in cognitive network science, text analysis and psychometrics. We introduce a quantitative framework measuring to what extent authors of suicide notes tend to associate emotional words in ways similar to how 200 individuals with high/low levels of depression, anxiety, and stress (on a DASS-21 scale) recall concepts together. In Study 1, we use text analysis to build one co-occurrence network out of suicide notes' texts and six fluency co-occurrence networks out of fluency data from individuals high/low on anxiety/stress/depression levels. We find that emotional word associations in suicide notes reflect mostly conceptual associations produced by low-anxiety individuals. In Study 2, we perform residual emotional profiling, i.e. measuring the emotional intensity of words remaining after removing from fluency networks those words mentioned and connected in the suicide notes' network. We find that only the high-anxiety fluency network displays negligible emotional residuals, indicating the highest overlap - in terms of emotions - between suicide notes' content and the concepts recalled by individuals with high anxiety levels. We discuss how these findings relate to existing psychological literature, along with their potential implications for understanding and quantifying suicidal behavior.
- Open Access
- 8 December 2022
- Published by Frontiers Media SA in Frontiers in Psychology
- Vol. 13, 917630
Cognitive networks detect structural patterns and emotional complexity in suicide notes
Communicating one's mindset means transmitting complex relationships between concepts and emotions. Using network science and word co-occurrences, we reconstruct conceptual associations as communicated in 139 genuine suicide notes, i.e., notes left by individuals who took their lives. We find that, despite their negative context, suicide notes are surprisingly positively valenced. Through emotional profiling, their ending statements are found to be markedly more emotional than their main body: The ending sentences in suicide notes elicit deeper fear/sadness but also stronger joy/trust and anticipation than the main body. Furthermore, by using data from the Emotional Recall Task, we model emotional transitions within these notes as co-occurrence networks and compare their structure against emotional recalls from mentally healthy individuals. Supported by psychological literature, we introduce emotional complexity as an affective analog of structural balance theory, measuring how elementary cycles (closed triads) of emotion co-occurrences mix positive, negative and neutral states in narratives and recollections. At the group level, authors of suicide narratives display a higher complexity than healthy individuals, i.e., lower levels of coherently valenced emotional states in triads. An entropy measure identified a similar tendency for suicide notes to shift more frequently between contrasting emotional states. Both the groups of authors of suicide notes and healthy individuals exhibit less complexity than random expectation. Our results demonstrate that suicide notes possess highly structured and contrastive narratives of emotions, more complex than expected by null models and healthy populations.
- 8 December 2022
- Published by Frontiers Media SA in Frontiers in Psychology
- Vol. 13, 917630
Open AccessCognitive networks detect structural patterns and emotional complexity in suicide notes
Communicating one's mindset means transmitting complex relationships between concepts and emotions. Using network science and word co-occurrences, we reconstruct conceptual associations as communicated in 139 genuine suicide notes, i.e., notes left by individuals who took their lives. We find that, despite their negative context, suicide notes are surprisingly positively valenced. Through emotional profiling, their ending statements are found to be markedly more emotional than their main body: The ending sentences in suicide notes elicit deeper fear/sadness but also stronger joy/trust and anticipation than the main body. Furthermore, by using data from the Emotional Recall Task, we model emotional transitions within these notes as co-occurrence networks and compare their structure against emotional recalls from mentally healthy individuals. Supported by psychological literature, we introduce emotional complexity as an affective analog of structural balance theory, measuring how elementary cycles (closed triads) of emotion co-occurrences mix positive, negative and neutral states in narratives and recollections. At the group level, authors of suicide narratives display a higher complexity than healthy individuals, i.e., lower levels of coherently valenced emotional states in triads. An entropy measure identified a similar tendency for suicide notes to shift more frequently between contrasting emotional states. Both the groups of authors of suicide notes and healthy individuals exhibit less complexity than random expectation. Our results demonstrate that suicide notes possess highly structured and contrastive narratives of emotions, more complex than expected by null models and healthy populations.
- Open Access
- 7 December 2022
- Published by Center for Open Science in OSF Preprints
The Struggle for Life among Words: How Cognitive Selection Shape Language Evolution
Like biological species, competition for survival is a constant among words in language. Previously, it has been shown that language evolves in response to cognitive constraints and over time becomes more learnable. Here, we use two complementary research paradigms to demonstrate how survival of words can be predicted by psycholinguistic properties that impact language production. In the first study, we analyzed survival of words in the context of inter-personal communication. We used a large-scale serial reproduction experiment in which stories were passed down along a transmission chain over multiple participants. The result shows that words that are acquired earlier in life, more concrete, more arousing, and more emotional have better chance to survive retellings. We reason that the same trend might scale up to language evolution over multiple generations of natural language users. If that is correct, the same set of psycholinguistics properties should also account for the change of word frequency in natural language corpora over evolutionary time. That is what we found in two large historical language corpora (study 2): early acquisition, concreteness and high arousal all predicts increasing word frequency over the past 200 years. By bridging micro-level behavioral preferences and macro-level language patterns, our study sheds light on the evolutionary mechanism of word competition.
- 7 December 2022
- Published by Center for Open Science in OSF Preprints
Open AccessThe Struggle for Life among Words: How Cognitive Selection Shape Language Evolution
Like biological species, competition for survival is a constant among words in language. Previously, it has been shown that language evolves in response to cognitive constraints and over time becomes more learnable. Here, we use two complementary research paradigms to demonstrate how survival of words can be predicted by psycholinguistic properties that impact language production. In the first study, we analyzed survival of words in the context of inter-personal communication. We used a large-scale serial reproduction experiment in which stories were passed down along a transmission chain over multiple participants. The result shows that words that are acquired earlier in life, more concrete, more arousing, and more emotional have better chance to survive retellings. We reason that the same trend might scale up to language evolution over multiple generations of natural language users. If that is correct, the same set of psycholinguistics properties should also account for the change of word frequency in natural language corpora over evolutionary time. That is what we found in two large historical language corpora (study 2): early acquisition, concreteness and high arousal all predicts increasing word frequency over the past 200 years. By bridging micro-level behavioral preferences and macro-level language patterns, our study sheds light on the evolutionary mechanism of word competition.
- Open Access
- 29 June 2022
- Published by Springer Nature in Memory & Cognition
- Vol. 50 (6) , 1284-1298
Diachronic semantic change in language is constrained by how people use and learn language
While it has long been understood that the human mind evolved to learn language, recent studies have begun to ask the inverted question: How has language evolved under the cognitive constraints of its users and become more learnable over time? In this paper, we explored how the semantic change of English words is shaped by the way humans acquire and process language. In Study 1, we quantified the extent of semantic change over the past 200 years and found that meaning change is more likely for words that are acquired later in life and are more difficult to process. We argue that it is human cognition that constrains the semantic evolution of words, rather than the other way around, because historical meanings of words were not easily accessible to people living today, and therefore could not have directly influenced how they learn and process language. In Study 2, we went further to show that semantic change, while bringing the benefit of meeting communicative needs, is cognitively costly for those who were born early enough to experience the change: Semantic change between 1970 and 2000 hindered processing speeds among middle-aged adults (ages 45–55) but not in younger adults (ages <25) in a semantic decision task. This hampering effect may have, in turn, curbed the rate of semantic change so that language does not change too fast for the human mind to catch up. Taken together, our research demonstrates that semantic change is shaped by processing and acquisition patterns across generations of language users.
- 29 June 2022
- Published by Springer Nature in Memory & Cognition
- Vol. 50 (6) , 1284-1298
Open AccessDiachronic semantic change in language is constrained by how people use and learn language
While it has long been understood that the human mind evolved to learn language, recent studies have begun to ask the inverted question: How has language evolved under the cognitive constraints of its users and become more learnable over time? In this paper, we explored how the semantic change of English words is shaped by the way humans acquire and process language. In Study 1, we quantified the extent of semantic change over the past 200 years and found that meaning change is more likely for words that are acquired later in life and are more difficult to process. We argue that it is human cognition that constrains the semantic evolution of words, rather than the other way around, because historical meanings of words were not easily accessible to people living today, and therefore could not have directly influenced how they learn and process language. In Study 2, we went further to show that semantic change, while bringing the benefit of meeting communicative needs, is cognitively costly for those who were born early enough to experience the change: Semantic change between 1970 and 2000 hindered processing speeds among middle-aged adults (ages 45–55) but not in younger adults (ages <25) in a semantic decision task. This hampering effect may have, in turn, curbed the rate of semantic change so that language does not change too fast for the human mind to catch up. Taken together, our research demonstrates that semantic change is shaped by processing and acquisition patterns across generations of language users.
- Open Access
- 13 December 2021
- Published by MDPI AG in Big Data and Cognitive Computing
- Vol. 5 (4) , 77
DASentimental: Detecting Depression, Anxiety, and Stress in Texts via Emotional Recall, Cognitive Networks, and Machine Learning
Most current affect scales and sentiment analysis on written text focus on quantifying valence/sentiment, the primary dimension of emotion. Distinguishing broader, more complex negative emotions of similar valence is key to evaluating mental health. We propose a semi-supervised machine learning model, DASentimental, to extract depression, anxiety, and stress from written text. We trained DASentimental to identify how N = 200 sequences of recalled emotional words correlate with recallers’ depression, anxiety, and stress from the Depression Anxiety Stress Scale (DASS-21). Using cognitive network science, we modeled every recall list as a bag-of-words (BOW) vector and as a walk over a network representation of semantic memory—in this case, free associations. This weights BOW entries according to their centrality (degree) in semantic memory and informs recalls using semantic network distances, thus embedding recalls in a cognitive representation. This embedding translated into state-of-the-art, cross-validated predictions for depression (R = 0.7), anxiety (R = 0.44), and stress (R = 0.52), equivalent to previous results employing additional human data. Powered by a multilayer perceptron neural network, DASentimental opens the door to probing the semantic organizations of emotional distress. We found that semantic distances between recalls (i.e., walk coverage), was key for estimating depression levels but redundant for anxiety and stress levels. Semantic distances from “fear” boosted anxiety predictions but were redundant when the “sad–happy” dyad was considered. We applied DASentimental to a clinical dataset of 142 suicide notes and found that the predicted depression and anxiety levels (high/low) corresponded to differences in valence and arousal as expected from a circumplex model of affect. We discuss key directions for future research enabled by artificial intelligence detecting stress, anxiety, and depression in texts.
- 13 December 2021
- Published by MDPI AG in Big Data and Cognitive Computing
- Vol. 5 (4) , 77
Open AccessDASentimental: Detecting Depression, Anxiety, and Stress in Texts via Emotional Recall, Cognitive Networks, and Machine Learning
Most current affect scales and sentiment analysis on written text focus on quantifying valence/sentiment, the primary dimension of emotion. Distinguishing broader, more complex negative emotions of similar valence is key to evaluating mental health. We propose a semi-supervised machine learning model, DASentimental, to extract depression, anxiety, and stress from written text. We trained DASentimental to identify how N = 200 sequences of recalled emotional words correlate with recallers’ depression, anxiety, and stress from the Depression Anxiety Stress Scale (DASS-21). Using cognitive network science, we modeled every recall list as a bag-of-words (BOW) vector and as a walk over a network representation of semantic memory—in this case, free associations. This weights BOW entries according to their centrality (degree) in semantic memory and informs recalls using semantic network distances, thus embedding recalls in a cognitive representation. This embedding translated into state-of-the-art, cross-validated predictions for depression (R = 0.7), anxiety (R = 0.44), and stress (R = 0.52), equivalent to previous results employing additional human data. Powered by a multilayer perceptron neural network, DASentimental opens the door to probing the semantic organizations of emotional distress. We found that semantic distances between recalls (i.e., walk coverage), was key for estimating depression levels but redundant for anxiety and stress levels. Semantic distances from “fear” boosted anxiety predictions but were redundant when the “sad–happy” dyad was considered. We applied DASentimental to a clinical dataset of 142 suicide notes and found that the predicted depression and anxiety levels (high/low) corresponded to differences in valence and arousal as expected from a circumplex model of affect. We discuss key directions for future research enabled by artificial intelligence detecting stress, anxiety, and depression in texts.