Current Issue

25 March 2026, Volume 58 Issue 3 Previous Issue    Next Issue
---
Reports of Empirical Studies


For Selected:

Download Citations EndNote Ris BibTeX

Toggle Thumbnails

Select

From overt deterrence to covert internalization: Moral effects of AI regulation and the moderating role of personality traits WANG Jianshu, JIANG Xiaowei, CHEN Yanan, WANG Minghui, DU Feng 2026, 58 (3):  381-398.  doi: 10.3724/SP.J.1041.2026.0381 Abstract ( 33 )   PDF (317KB) ( 10 )   As generative artificial intelligence (GenAI) evolves into social agents with autonomous influence, its impact on human moral decision-making is becoming increasingly significant. Current regulatory models are often grounded in the "rational person hypothesis, " which assumes uniform responses to ethical constraints. This perspective, however, overlooks the profound moderating role of personality traits in moral choices, leading to divergent regulatory effects and a loss of efficiency. The Dark Triad of personality (narcissism, Machiavellianism, and psychopathy) is a robust predictor of moral deviation. To address this gap, we constructed a "Regulation Type × Personality Trait" interaction model. We hypothesized that the effectiveness of different AI-driven intervention strategies—namely explicit regulation, implicit incentives, and moral feedback—would be significantly moderated by individuals' Dark Triad traits when making decisions about honesty. A series of experiments were conducted to test our hypotheses. The study utilized a modified coin-flip task where participants privately guessed and reported outcomes, a paradigm designed to create opportunities for dishonest behavior for personal gain. Participants' honesty rates and reaction times were recorded as the primary dependent variables. Across the experiments, we manipulated the AI-driven intervention strategies. These strategies included: (1) explicit (visible) versus implicit (invisible) AI surveillance which involved potential penalties for dishonesty; (2) implicit monetary incentives which rewarded consistent honesty; and (3) moral feedback which provided textual messages in response to honest or dishonest reports. Prior to the behavioral tasks, participants' personality traits were measured using the validated Short Dark Triad (SD3) scale. The results supported our hypotheses, demonstrating significant interactions between intervention types and personality traits. In Experiment 1, explicit AI surveillance significantly increased honest reporting (t(45) = 4.59, p < 0.001), particularly among individuals with high levels of Machiavellianism (t(25) = 4.60, p = 0.005) and psychopathy (t(28) = 4.44, p < 0.001). In Experiment 2, invisible AI surveillance also enhanced honesty but was less effective than visible AI surveillance, F(2, 90) = 18.10, p < 0.001. Notably, invisible surveillance resulted in the shortest reaction time (RT = 0.49), F(2, 90) = 34.10, p < 0.001. High Machiavellian participants displayed greater honesty under visible surveillance (OR = 0.70, p = 0.013) but were more dishonest without or under invisible AI surveillance. In Experiment 3, potential financial rewards increased reaction time (F(2, 118) = 58.59, p < 0.001), while high Machiavellian individuals showed reduced honesty during the internalization stage, t(57.98) = -2.04, p = 0.044. In Experiment 3a and 3b, financial incentives promoted honesty more effectively than moral messaging during the reward stage (t(120) = 3.07, p = 0.003) and maintained this effect into the internalization stage (t(120) = 2.06, p = 0.041), demonstrating the robustness of monetary influence. High Machiavellian participants sustained higher honesty levels in the internalization stage (OR = 1.96, p < 0.001). In contrast, narcissistic participants showed resistance to moral messaging, especially during the reward stage, t(49.95) = -2.55, p = 0.013. This study was the first to systematically reveal the critical moderating role of the Dark Triad personality traits in AI ethical regulation. The findings challenge the traditional 'rational person' paradigm by empirically demonstrating the significant personality-based heterogeneity of regulatory effects. The core contribution of this research is the proposal of an innovative concept: 'personality-regulated regulation.' This framework provides a vital theoretical and practical foundation for designing future AI ethical intervention strategies that are contextualized and personalized. Such an approach allows for the optimization of regulatory resource allocation and enhances overall regulatory efficacy, moving beyond one-size-fits-all models. References | Related Articles | Metrics

Select

LLMs amplify gendered empathy stereotypes and influence major and career recommendations DAI Yiqing, MA Xinming, WU Zhen 2026, 58 (3):  399-415.  doi: 10.3724/SP.J.1041.2026.0399 Abstract ( 22 )   PDF (1257KB) ( 8 )   As large language models (LLMs) are increasingly deployed in sensitive domains such as education and career guidance, concerns have grown about their potential to amplify gender bias. Prior research has documented occupational gender stereotypes in LLMs, such as associating men with technical roles and women with caregiving roles However, less attention has been paid to whether these models also encode deeper socio-emotional traits in gender-based ways. A persistent societal stereotype holds that “women are more empathetic than men”, a belief that can shape career expectations. This study investigated whether LLMs reflect or even exaggerate gender stereotypes related to empathy and examined the contextual factors (e.g., input language, gender-identity priming) that might influence the expression of these stereotypes. We hypothesized that LLMs would exhibit stronger gendered empathy stereotypes than human participants, that these biases would vary according to linguistic and social cues in prompts; and that these stereotypes would manifest in real-world major/career recommendation scenarios. We conducted three studies to test these hypotheses. Study 1 compared judgments about empathy from human participants (N = 626) with those generated by six leading LLMs (GPT-4o, GPT-4-Turbo, GPT-3.5-Turbo, DeepSeek-reasoner, DeepSeek-chat, ERNIE-Bot). Twelve story-based scenarios, adapted from the Empathy Questionnaire, covered emotional empathy, attention to others’ feelings, and behavioral empathy. For each scenario, participants and LLMs inferred the protagonist’s gender based on their empathetic behavior. Study 2 examined how manipulating input language (English vs. Chinese) and gender-identity priming (male vs. female) influenced the expression of these stereotypes. Study 3 extended this paradigm to a real-world application: we prompted LLMs to recommend 16 pre-selected university majors and 16 professions (categorized into high- and low-empathy groups) to individuals of different genders, requesting explanatory rationales for each recommendation. Results indicated that LLMs displayed significantly stronger gendered empathy stereotypes than human participants (Study 1). English prompts and female priming elicited stronger “women = high empathy, men = low empathy” associations (Study 2). In the recommendation tasks, LLMs more often suggested high-empathy majors and professions (e.g., nursing, education, psychology) for women, and low-empathy, STEM-related fields for men (Study 3). Together, these findings suggest that LLMs not only internalize gendered empathy stereotypes but also express them in context-dependent ways, producing measurable downstream effects in applied decision-making tasks. =Overall, our findings underscore the need for critical evaluation of how LLMs represent and amplify social stereotypes, especially in socio-emotional domains such as empathy. This research contributes to understanding the sources of AI bias by showing that LLMs may exaggerate gender norms beyond human levels. It also highlights the complex interplay between language and gender identity in shaping algorithmic behavior. Practically, the results raise important ethical concerns about fairness in AI-driven decision-making systems and highlight the urgency of developing more robust bias-mitigation strategies in multilingual contexts. References | Related Articles | Metrics

Select

Large language models capable of distinguishing between single and repeated gambles: Understanding and intervening in risky choice ZHOU Lei, LI Litong, WANG Xu, OU Huafeng, HU Qianyu, LI Aimei, GU Chenyan 2026, 58 (3):  416-436.  doi: 10.3724/SP.J.1041.2026.0416 Abstract ( 12 )   PDF (3052KB) ( 4 )   Risky choice (RC) is a common and important form of decision making in daily life. Its theoretical development primarily follows two major theories: normative theory and descriptive theory. The paradigms of single- and repeated-play gambles can provide an effective framework for distinguishing between the theories. However, prior research lacks direct observations of the decision-making process, which can limit the deep understanding of individual behaviour and hinder the development of effective behavioural interventions. In recent years, large language models (LLMs) have demonstrated highly human-like characteristics by not only simulating human preferences in behavioural performance but also exhibiting similar reasoning pathways. This offers a promising solution to the aforementioned limitations. This study, which is grounded in the classic RC paradigms of single versus repeated gambles, investigates the capability of LLMs to simulate and understand risk preferences and decision-making processes. Specifically, this study explores the potential of LLMs’ understanding of decision strategies to generate intervention texts and evaluates their effectiveness in influencing human decisions. This work comprises three studies. In Study 1, GPT-3.5 and GPT-4 were employed to simulate human responses to gambling decisions under nine probability conditions (with constant expected value), which generated a total of 3, 600 responses across single and repeated gamble scenarios. In Study 2, LLM-generated strategies were constructed through a three-stage process (decision rationale extraction, strategy generation and quality evaluation), then the human participants were required to complete decision-making tasks in two experiments: Experiment 1 replicated the medical/financial scenarios (N = 349, N male = 174, M age = 21.79) of Sun et al. (2014) in a 2 (context: medical vs. financial) × 2 (application frequency: single vs. repeated) within-subjects design, and Experiment 2 examined digital contexts with a 2 (context: content creation vs. e-commerce marketing) × 2 (frequency: single vs. repeated) mixed design (context as between subjects). Subsequently, DeepSeek-R1 was used to perform the same tasks and generate strategy texts through the three-stage process. Finally, the participants were instructed to evaluate their acceptance of the LLM-generated strategies. Study 3 extended the Study 2 methodology to determine whether the LLM-generated intervention texts could reverse the participants’ classic choice preference across the single versus repeated gamble scenarios. The Study 2 experimental contexts (Experiment 1: medical vs. financial, N = 460, N male = 205, M age = 21.80; Experiment 2: content creation vs. e-commerce marketing, N = 240, N male = 106, M age = 29.12) were mirrored in Study 3, in which strategically designed intervention texts were presented during the decision-making tasks to test their capacity to modify the participants’ inherent risk preference between the single and repeated gamble conditions and evaluate the persuasive efficacy of LLM-generated strategies on human decision biases. Study 1 shows that the LLMs (GPT-3.5 and GPT-4) can successfully replicate the typical human pattern of risk aversion in single-play scenarios and risk seeking in repeated-play scenarios, though both models demonstrated an overall stronger tendency toward risk seeking compared with the human participants. Study 2 demonstrates that the human participants preferred low-EV certain options in single-play contexts and high-EV risky options in repeated-play contexts in both experiments. The participants also showed high agreement with the strategies generated by the LLMs in different scenarios. Study 3 confirms that the LLM-generated intervention texts can significantly influence the participants’ choice tendency in all four scenarios, with strong intervention effects observed in the single-play contexts. The LLM intervention strategies are characterised by reliance on expected value computations (normative) when promoting RCs and emphasis on certainty and robustness (descriptive) when promoting safe choices. In summary, this study demonstrates that (1) LLMs can effectively simulate context-dependent human preferences in RC, particularly the shift from risk aversion in single plays to risk seeking in repeated plays; (2) LLMs can distinguish between the logic underlying single and repeated gambles and apply normative and descriptive reasoning accordingly to externalise decision strategies; and (3) the decision strategies extracted from LLM-generated reasoning can be used to construct effective intervention texts that can alter human preferences in classic risk decision tasks, thereby validating the feasibility and effectiveness of an LLM-based cognitive intervention pathway. This study offers a new technological paradigm for AI-assisted decision intervention and expands the application boundary of LLMs to human cognitive process modelling and regulation. References | Related Articles | Metrics

Reports of Empirical Studies

Select

Cognitive neural mechanisms of tonal patterns and semantic processing in poetry ZHANG Jingjing, SHI Ying, DENG Shanwen, LI Jiabin, CHEN Qingrong 2026, 58 (3):  437-449.  doi: 10.3724/SP.J.1041.2026.0437 Abstract ( 25 )   PDF (690KB) ( 4 )   Music has long been considered the soul of poetry, a reflection of their shared origins across diverse cultural contexts. Historically, poetry and music were inseparable, with poems frequently performed through songs. As the arts evolved, however, poetry gradually shifted toward a focus on literal meaning, shedding its musical accompaniment and entering an era of “words without tunes.” Yet the musical essence of poetry endures, embedded in the rhythmic, tonal, and melodic qualities of language itself. In classic Chinese poetry, tonal patterns serve as a defining prosodic feature, shaping both the rhythmic flow and the musical character of a poem. The present study investigated the neural mechanisms underlying tonal patterns in classic Chinese poetry and their interaction with semantic processing. Forty-eight participants with extensive experience reading classic Chinese poems took part in an electroencephalography (EEG) experiment. They were presented with four versions of classic Chinese poems in a pseudo-random order: (1) semantically congruous with regular tonal patterns (S+T+), (2) semantically congruous with irregular tonal patterns (S+T-), (3) semantically incongruous with regular tonal patterns (S-T+), and (4) semantically incongruous with irregular tonal patterns (S-T-). Each participant viewed 100 experimental stimuli and additional 40 filler stimuli. After reading each poem, participants rated the reasonableness of both its semantics and tonal patterns. Throughout the experiment, both behavioral responses and EEG data were recorded. Accuracy rates and event-related potentials (ERP) were analyzed using linear mixed-effects models via the lmerTest and emmeans packages in the R, and EEG preprocessing was conducted using the MNE software package in Python. Accuracy results showed that participants performed best on the S+T+ condition, whereas congruous poems with irregular tonal patterns (S+T-) yielded the lowest accuracy. ERP analyses revealed that tonal patterns exerted a continuous influence on both early and late stages of poetry reading. Specifically, in the P200 window, an interaction between tonal patterns and scalp region indicated that irregular tonal patterns elicited larger P200 amplitudes at anterior and central sites. In the N400 time window, a significant interaction between semantics and tonal patterns emerged: semantically incongruous lines produced a more pronounced negative component than congruous lines when tonal patterns were regular, whereas this semantic difference diminished under irregular patterns. Furthermore, in the LPC time window, irregular tonal patterns evoked larger positivities for semantically incongruous poems, while semantic congruity eliminated the tonal pattern effect altogether. Finally, deep learning models trained on the EEG data reliably distinguished among the four experimental conditions, indicating robust neural signatures associated with the combined tonal-semantic processing. =In summary, the current findings underscore the significant impact of tonal patterns on the reading of classic Chinese poetry. Irregular patterns modulated phonological representation at early stages and constrained semantic comprehension at later stages. Supporting the neurocognitive poetics model (NCPM) of literary reading, these results shed light on how prosodic elements and semantic meaning dynamically interact over time during poetic processing. References | Related Articles | Metrics

Select

Modulation of rhythmic temporal attention by conscious awareness: Evidence from behavior, hierarchical drift-diffusion modeling, and EEG measures LIANG Xingjie, CHEN Huifang, WANG Luyao, SUN Yanliang 2026, 58 (3):  450-466.  doi: 10.3724/SP.J.1041.2026.0450 Abstract ( 16 )   PDF (1235KB) ( 6 )   Temporal cues enable individuals to anticipate upcoming events, thereby facilitating goal-directed behavior. While temporal association cues are known to engage endogenous temporal attention that is modulated by conscious perception, it remains unclear whether rhythmic cues—typically considered to evoke exogenous temporal attention—are similarly affected by the state of consciousness. Addressing this gap, the present study investigated whether rhythmic temporal attention is subject to modulation by conscious awareness and whether it involves endogenous cognitive components akin to those recruited by symbolic temporal cues. Two experiments were conducted, each involving 24 different Chinese participants and comprising three conditions: (a) a rhythmic cueing task under conscious perception, (b) the same task under unconscious perception manipulated via high-frequency flicker (50 Hz), and (c) a two-alternative forced-choice awareness check. Experiment 2 replicated the design of Experiment 1 with simultaneous EEG recordings. Participants performed an orientation discrimination task in rhythmic versus random cue conditions, with inter-stimulus intervals (ISIs) of either 800 ms or 1300 ms to compare sub-second and supra-second timing. Behavioral results showed robust temporal attention effects in both conscious and unconscious states, though significantly larger under conscious perception. ERP analyses revealed that rhythmic cues elicited greater contingent negative variation (CNV) amplitudes when participants were conscious, indicating enhanced temporal preparation at the neural level. Hierarchical drift-diffusion modeling (HDDM) further showed that under conscious perception, rhythmic cues reduced decision boundaries, suggesting more confident and efficient decision-making—a hallmark of endogenous control. These effects were absent under unconscious conditions. Additionally, faster responses in supra-second versus sub-second intervals support the segmented timing hypothesis and indicate that longer temporal contexts may recruit higher-order cognitive processes. Importantly, time-frequency analysis revealed stronger alpha-band (8~12 Hz) suppression during the rhythmic encoding phase under conscious perception, particularly over frontal and occipital regions, with wider spatial distribution in the supra-second interval. This enhanced alpha desynchronization suggests greater attentional engagement and top-down modulation of sensory areas, supporting the notion that conscious perception of rhythmic structure facilitates the neural entrainment of anticipatory attention. Together, these findings challenge the view that rhythmic temporal attention is purely exogenous, showing instead that it contains an endogenous component that is modulated by the state of consciousness. This study provides converging behavioral, electrophysiological, and computational evidence for a dual-process account of rhythmic temporal attention and offers novel insights into the interaction between temporal structure and awareness in shaping anticipatory cognition. References | Related Articles | Metrics

Select

Behavioral theta oscillations in cross-modal stimulus conflict and response conflict processing XU Honghui, XU Yiran, YANG Guochun, NAN Weizhi, LIU Xun 2026, 58 (3):  467-479.  doi: 10.3724/SP.J.1041.2026.0467 Abstract ( 17 )   PDF (935KB) ( 5 )   Theta oscillations are closely associated with cognitive control. Accumulating evidence indicates their involvement in processing both cross-modal stimulus conflict and response conflict. However, the relationship between theta oscillations and the magnitude of these conflicts remains unclear. Research on behavioral oscillations, which shows that theta rhythms are linked to periodic patterns in behavioral performance, provides a novel perspective for investigating how theta oscillations are related to the magnitude of cross-modal stimulus conflict and response conflict. To address this issue, we used an audiovisual Stroop task with a 2-to-1 stimulus-response mapping and a time-resolved behavioral approach. Given previous evidence that sensory modality modulates theta oscillations during cross-modal stimulus conflict and response conflict processing, we designed two distinct tasks. In Experiment 1 (N = 43), participants responded to auditory stimuli while ignoring visual distractors (auditory task). In Experiment 2 (N = 40), participants responded to visual stimuli while ignoring auditory distractors (visual task). This design allowed us to investigate the relationship between theta oscillations and the magnitude of cross-modal stimulus and response conflicts in the auditory and visual tasks, respectively. The results demonstrated that the rhythmic processing of task-relevant stimuli was modulated by task-irrelevant stimuli. When the task-irrelevant stimuli were either the same as or different from the task-relevant stimuli (including both stimulus and response incongruency), response times exhibited rhythmic fluctuations in the theta band (4~7.5 Hz). In contrast, when task-irrelevant stimuli were neutral, the processing was characterized by oscillations in the alpha band (9.4~10 Hz). Furthermore, we found that the magnitude of the cross-modal response conflict itself fluctuated rhythmically at a theta frequency (3.8 Hz) in the auditory task, whereas the magnitude of the cross-modal stimulus conflict fluctuated rhythmically at a theta frequency (5.6 Hz) in the visual task. In summary, the present study demonstrates that cognitive control processes exhibit theta oscillations at the behavioral level, which directly modulate the magnitude of cross-modal stimulus conflict and response conflict. These findings elucidate the relationship between theta oscillations and conflict magnitude, and extend the rhythmic theory of attention to the domain of cognitive control in conflict processing. References | Related Articles | Metrics

Select

Buildup of a self-compassion sentence collection in the construct of implementation intention and its application in emotion regulation SONG Yi, FU Xiaotong, YUAN Jiajin, SUN Meng, YANG Jiemin 2026, 58 (3):  480-499.  doi: 10.3724/SP.J.1041.2026.0480 Abstract ( 32 )   PDF (1048KB) ( 17 )   Emotional situations in real life are highly diverse, yet existing emotion regulation training approaches often lack adaptivity to such diversity. Self-compassion, an emotion regulation strategy characterized by kindness and non-judgmental understanding toward one's own suffering, has demonstrated emotional benefits. However, its effectiveness is constrained by cognitive demands, limiting its applicability among individuals with emotional disorders or cognitive impairments. Implementation intention, specifying how to response in a given situation based on the goal with a typical structure as “if + situation, then + reaction”, can effectively reduce the cognitive demands of self-regulation and enhance its regulatory effect. To improve the emotional regulation efficacy of self-compassion, this study integrated implementation intention with self-compassion to establish a standardized sentence collection (Experiment 1) and evaluated its effectiveness in regulating negative emotions among college students with depressive tendencies (Experiment 2~3). In Experiment 1, situational and response sentences were prepared following the “if + situation, then + reaction” structure. Forty negative and forty neutral situational sentences were developed, along with 120 implementation intention-based self-compassion (II-SC) statements and 120 implementation intention-based neutral (II-N) statements. A total of 106 participants (age: 20.17±1.90 years; 66 females) evaluated these sentences on dimensions such as emotional valence, arousal, and relevance to self-compassion. By presenting negative situational sentences to induce negative emotions, followed by II-SC statements, a substantial increase in pleasure was observed, supporting the efficacy of II-SC in emotion regulation. Experiment 2 employed a 2 (measuring time: pre-test vs. post-test) × 3 (group: II-SC vs. TSC vs. Control) mixed factorial design to examine the effect of II-SC on negative emotion regulation. Ninety college students with depressive tendencies (age: 20.07±1.80 years; 82 females) were randomly assigned to one of the three groups (II-SC, TSC or Control group). After baseline assessments of emotional state, participants received either II-SC, TSC (traditional self-compassion) guidance, or no guidance (Control group). Negative emotions were then induced through negative self-evaluation sentences accompanied by sad music, followed by post-test on emotional measures. Building upon Experiment 2, Experiment 3 incorporated physiological indicators including heart rate and electrodermal activity to more comprehensively and objectively assess emotional responses. Sixty-nine college students with depressive tendencies (age: 20.01±1.38 years; 65 female) participated. In this experiment, negative emotions were elicited through challenging arithmetic tasks to test whether the II-SC sentences produced increased emotion regulation effects compared with TSC under conditions of cognitive resource depletion. Results indicate that (1) II-SC sentence collection can significantly alleviate negative emotions, and showed a satisfactory reliability and validity. (2) When negative emotions were induced by negative self-evaluation, both II-SC and TSC sentences effectively reduced negative affect, with no significant difference in efficacy; however II-SC required less cognitive effort. (3) When negative emotions were elicited through frustrating tasks that depleted cognitive resources, both II-SC and TSC alleviated negative affect, but the resource dependence of TSC reduced its effectiveness, resulting in better regulatory effects for II-SC. Nevertheless, no significant group differences were found in perceived regulation effort, and neither physiological activation associated with negative emotion induction. In conclusion, the current study developed a standardized and valid self-compassion sentence collection based on implementation intention, and demonstrated its effectiveness in regulating negative emotions. These II-SC sentences enable individuals with depressive tendencies to engage in automatic and efficient emotion regulation through self-compassion, thereby enhancing its regulatory efficacy and offering a potential approach for coping with the diverse stressors encountered in daily life. References | Related Articles | Metrics

Select

Neural mechanisms of binaural beats in pain modulation LI Xiaoyun, WU Qiqi, JIANG Liwen, LU Xuejing, PENG Weiwei 2026, 58 (3):  500-515.  doi: 10.3724/SP.J.1041.2026.0500 Abstract ( 9 )   PDF (762KB) ( 3 )   Non-pharmacological interventions for pain have been a crucial topic within psychology and neuroscience disciplines for many years. Among them, rhythmic auditory stimulation has gained increasing attention due to its non-invasive nature, repeatability, and suitability for daily use. Binaural beats (BBs), which comprise a virtual, rhythmic signal generated by presenting slightly different frequencies to each ear, can induce cortical oscillations at matching frequencies and thereby modulate brain activity. BBs engage more complex central auditory processing compared to monaural beats (MBs), which are physically mixed sounds that primarily act as external rhythmic stimuli. There is an inverse relationship between alpha activity and pain perception, and enhanced alpha oscillations have been proposed to play an analgesic role. Accordingly, this study employs alpha-band BBs as an experimental stimulus, with MBs and white noise as control conditions, to investigate their analgesic effects and underlying neural mechanisms. Using a within-subjects design with three auditory conditions (BBs, MBs, and white noise), we assessed subjective pain ratings (intensity and unpleasantness), laser-evoked potentials (LEPs), and spontaneous EEG dynamics. To comprehensively capture neural modulation, we combined spectral power analysis with EEG microstate analysis to examine the dynamic reorganization of brain networks during auditory stimulation with BBs. No significant differences in subjective pain ratings were observed across conditions. However, BBs were found to uniquely modulate brain dynamics. Both BBs and MBs significantly decreased gamma-band power during stimulation compared to white noise, indicating a similar effect between these rhythmic auditory inputs on high-frequency activity. Importantly, microstate analysis revealed BB-specific changes: BBs enhanced the occurrence of microstate A (associated with primary auditory processing) while reducing the presence of microstate C (linked to introspective and self-referential processing). Mediation analysis further showed that BBs indirectly modulated pain-related P2 amplitudes, indicative of attentional allocation to nociceptive stimuli, through reducing the transition probability between microstate C and microstate D (involved in attentional reorientation); these factors together may modulate subjective pain experiences. This dynamic pathway suggests that BBs can alter pain processing by reshaping functional brain states and modulating the deployment of attentional resources. In summary, while BBs did not produce robust behavioral analgesia, they produced significant neural modulatory effects—potentially by reducing dynamic switching between the default mode network and attention-related networks. Our methodology of EEG microstate analysis to investigate pain modulation by rhythmic auditory stimulation offers a novel perspective for evaluating non-pharmacological neuromodulation. Theoretically, our findings call for a shift from frequency-centric views toward a state-dependent framework emphasizing dynamic brain network reorganization. Future studies may explore personalized rhythmic stimulation protocols tailored to individual brain dynamics to enhance the clinical application of BBs in chronic pain and affective disorders. References | Related Articles | Metrics

Select

The development and motivations of children’s third-party intervention preference in group cooperation norm violation ZHU Naping, ZHANG Xia, ZHOU Jie, LI Yanfang 2026, 58 (3):  516-533.  doi: 10.3724/SP.J.1041.2026.0516 Abstract ( 18 )   PDF (991KB) ( 3 )   Third-party intervention plays a critical role in maintaining large-scale human group cooperation. Previous studies on children’s third-party intervention mainly focused on explicit norm violations, such as unfair distribution or damage to others’ belongings. Different from explicit violations, free-riding in group cooperation, where an individual benefits from others' contributions without incurring the corresponding costs) involves implicit norms and lacks clearly identifiable victims. Recognizing and addressing free riders in group cooperation from a third-party perspective is an important aspect of children’s social norms acquisition and the development of cooperative behavior. This study examined the development and motivations of third-party intervention preference among children aged 4 to 11 years old in scenarios involving group cooperation norms violations. =In Study 1, 141 children (70 boys, Age range: 4.06~11.96 years, Mage = 8.02, SD = 2.30) heard a cooperation story in which one group member shirked effort by playing football instead of helping clean the classroom but still shared in the reward. Children’s moral evaluation, anger toward free-riding behavior, and their intervention preferences were measured in sequence. In Study 2, 125 children (63 boys, Age range: 4.34~11.68 years, Mage = 8.01, SD = 2.31) completed a similar task involving materials contribution to a collective resource. To test cross-situational stability and rule out majority-influence effects, the group composition was adjusted to two cooperators and two free riders. In this story, a four-member group jointly dropped gold coins into a magic jar to get more gold coins. Among them, two members of the group each dropped one gold coin into the magic jar, while the other two members did not. In the end, the two coins that were dropped into the magic jar turned into four gold coins, and then each member of the group received one gold coin. Similarly to Study 1, children’s moral evaluation, anger feeling on free-riding behavior, intervention preferences and motivations were measured. Across both studies, children consistently evaluated free-riding negatively and reported anger toward it. As third-parties, children preferred intervention in free-riding behavior over non-intervention across all ages. With age, children’s preferences for third-party intervention showed a cross-situational stable developmental trend, gradually shifting from rewarding cooperators to punishing free riders. Importantly, the shift occurred earlier in effort-based scenarios (ages 5.51~5.67) than in material-based scenarios (ages 8.21~8.22). In terms of the motivations of intervention preferences, Study 2 found that the internal motivations for children to reward cooperators reflect both deontological motivation and consequentialist motivation. In contrast, the motivation to punish free-riders changes with age, gradually shifting toward the consequentialist motivation after age six. These findings indicated that the underlying motivations driving children’s intervention preferences are both specific and age-dependent. These findings demonstrate that even young children can morally evaluate norm violations in group cooperation and engage in third-party interventions. Their intervention preferences develop in a stable, cross-situational manner, gradually shifting from rewarding prosocial behavior to punishing norm violators. However, the motivations underlying reward and punishment are distinct and age-dependent. This research provides valuable insights into the development of children’s cooperative behavior and their understanding of group cooperation norm. References | Related Articles | Metrics

Select

The interactive effects of intrinsic and extrinsic rewards and context stability on volunteers’ volunteering habits QU Guoliang, JU Enxia, XUE Yining, CHEN Xuhai, LUO Yangmei 2026, 58 (3):  534-557.  doi: 10.3724/SP.J.1041.2026.0534 Abstract ( 29 )   PDF (762KB) ( 5 )   Volunteering is a typical form of prosocial behavior in which individuals devote their time, energy, and skills to benefit others and society. Volunteering plays a vital role in enhancing volunteers’ individual well-being and mental health, strengthening community governance, promoting social cohesion, fostering economic development, and advancing societal progress. Despite these well-documented benefits, China’s volunteering sector continues to face several practical challenges, including low public willingness to participate, limited sustainability of engagement, and high volunteer turnover rates. Thus, there is a crucial need to identify factors that can promote long-term and sustainable volunteer participation. Studies to date have primarily focused on enhancing volunteers’ motivation and role identity. In contrast, the role of habit formation in sustaining volunteering has received relatively little scholarly attention. Moreover, the relationships linking intrinsic and extrinsic rewards, context stability, and volunteering habits remain insufficiently understood. To address these gaps, three studies were conducted to systematically examine how intrinsic and extrinsic rewards and context stability influence volunteering habits. In Study 1, data were collected using semi-structured interviews with 25 volunteers who exhibited strong volunteering habits. Thematic analysis identified three types of intrinsic rewards (meaningfulness, happiness, and self-worth) and three types of extrinsic rewards (honorary rewards, material rewards, and social support and recognition). In addition, five forms of context stability (time, place, activity type, people, and mood) were identified as well. These factors may be potential predictors of volunteering habits. Study 2 built on these qualitative findings through the development of quantitative measures of intrinsic rewards, extrinsic rewards, and context stability. A total of 1, 572 community volunteers and 853 student volunteers were surveyed using these scales and the Self-Report Habit Index for Volunteering. Multiple linear regression analyses revealed that intrinsic rewards, extrinsic rewards, and context stability significantly and positively predicted the strength of volunteering habits. Furthermore, a significant interaction between intrinsic rewards and context stability was found, such that when context stability was low, the positive effect of intrinsic rewards on volunteering habits became stronger. In contrast, the interaction between extrinsic rewards and context stability was not significant. These patterns were consistent across both community and student volunteer samples. To complement the cross-sectional nature of Study 2, Study 3 adopted a longitudinal design in which three waves of data were collected, separated by three-month intervals, from 623 volunteers. Cross-lagged panel modeling revealed that intrinsic rewards at Time 1 (T1) significantly predicted volunteering habits at Time 2 (T2), whereas extrinsic rewards at T1 only marginally predicted volunteering habits at T2, and context stability at T1 did not significantly predict volunteering habits at T2. Similarly, intrinsic rewards at T2 significantly predicted volunteering habits at Time 3 (T3), whereas extrinsic rewards and context stability at T2 did not exhibit significant predictive effects on volunteering habits at T3. Furthermore, the interaction between intrinsic rewards and context stability at T2 was significant, while the interaction between extrinsic rewards and context stability at T2 was only marginally significant. Specifically, when context stability at T2 was low, the positive effects of either intrinsic or extrinsic rewards at T2 on volunteering habits at T3 became stronger. In conclusion, intrinsic rewards (meaningfulness, happiness, and self-worth), extrinsic rewards (honorary rewards, material rewards, and social support and recognition), and context stability (time, place, activity type, people, and mood stabilities) emerged as key factors that enhance the strength of volunteering habits. Of these factors, intrinsic rewards were the most stable and robust predictors. Moreover, both intrinsic and extrinsic rewards interacted with context stability to predict volunteering habits, indicating that rewards exert stronger effects when the volunteering context is less stable. These findings offer valuable practical implications for sustaining long-term volunteer engagement. For temporary or emergency volunteer activities (e.g., natural disaster relief and large-scale social events) when service time and location are unstable and consistent contextual cues are lacking, it becomes particularly important to provide diverse forms of intrinsic and extrinsic rewards (such as honorary recognition, material subsidies, or social acknowledgment). Such incentives can enhance volunteers’ willingness to participate, encourage their repeated engagement, and ultimately facilitate the formation of volunteering habits. At the theoretical level, these results extend habit theory to the volunteering domain and offer new insights into how reward structures and contextual features can jointly facilitate the maintenance of prosocial behaviors. References | Related Articles | Metrics

Select

Factor retention in exploratory factor analysis based on LSTM GUO Lei, QIN Haijiang 2026, 58 (3):  558-568.  doi: 10.3724/SP.J.1041.2026.0558 Abstract ( 21 )   PDF (1150KB) ( 4 )   Psychological research focuses on the latent traits of individuals, necessitating clear operational definitions to delineate the constructs of interest. Following this, the exploration and description of the dimensions and characteristics of these traits are essential. Exploratory Factor Analysis (EFA) is a pivotal statistical method for identifying these latent dimensions, widely utilized, especially in the development of psychological scales and instruments. A critical aspect of employing EFA is the accurate determination of the number of factors. Underestimating the number of factors may result in the omission of theoretically significant psychological structures or sub-dimensions, leading to the loss of critical information, increased estimation errors in factor loadings, and diminished accuracy of factor scores. Conversely, overestimating the number of factors may lead to factors splitting, where the primary loadings of manifest variables are dispersed across multiple factors, thereby weakening the association between the manifest variables and the intended factor. Moreover, this may result in a model characterized by undue complexity and structures of limited practical or theoretical utility. To address these challenges, researchers have proposed various methods, including the Kaiser criterion (i.e., eigenvalues greater than one), the empirical Kaiser criterion, Parallel Analysis, the Hull method, Comparison Data, Factor Forest, and Comparison Data Forest. With the rapid advancement of machine learning, its application in EFA has begun to attract attention. This study introduces an innovative approach by treating eigenvalues as sequential data and leveraging Long Short-Term Memory (LSTM) networks to construct a predictive model. The performance of the LSTM-based method was subsequently evaluated through extensive simulations and empirical studies under diverse data conditions, demonstrating its robustness and applicability. The findings of the study indicate that: (1) After hyperparameter tuning, an optimal combination was identified, enabling the LSTM model to achieve excellent performance across accuracy, precision, and other evaluation metrics, demonstrating high classification capability. (2) In the simulation study, the LSTM model significantly outperformed Comparison Data Forest, the Empirical Kaiser Criterion, and Parallel Analysis under nearly all data conditions, with an average improvement in estimation accuracy of 48.50% and a maximum improvement of 171.09%. Furthermore, an empirical study was conducted using data from a parental psychological control scale administered to a cohort of 987 high school students in a city in 2022. Both traditional methods and the LSTM approach were employed to assess ecological validity. The results demonstrated that the LSTM provided the most accurate estimation of the number of factors, while the CDF method exhibited a significant tendency to overestimate. Overall, the LSTM proposed in this study demonstrates strong practical value and is worthy of broader adoption. Researchers can use the R package LSTMfactors to call the LSTM trained in this study to analyze empirical data. References | Related Articles | Metrics