Loading...
  Office Online
    Online Submission
    Office Work
    Peer Review
    Editor Work
    Editor-in-chief
  Journal Online
    Forthcoming Articles
    Current Issue
    Advanced Search
    Archive
    TOP Read
    TOP Download
    Email Alert
    
  • Table of Content
       , Volume 48 Issue 3 Previous Issue    Next Issue
    For Selected: View Abstracts Toggle Thumbnails
    Influence of selective attention on implicit learning with auditory stimuli
    LI Xiujun; SHI Wendian
    Acta Psychologica Sinica. 2016, 48 (3): 221-229.   DOI: 10.3724/SP.J.1041.2016.00221
    Abstract   PDF (425KB) ( 6182 )

     

    Implicit learning refers to people’s tendency to acquire complex regularities or patterns without intention or awareness (Reber, 1989). Given regularities are acquired without intention, and largely unconsciousness, implicit learning is often considered to occur without attention. The processes responsible for such learning were once contrasted with a selective intentional “system” (Guo et al., 2013; Jiang & Leung, 2005). However, more recent researches show that actually implicit learning processes are highly selective (Eitam, schul, & Hassin, 2009; Eitam et al., 2013; Tanaka, Kiyokawa, Yamada, Dienes, & Shigemasu, 2008; Weiermann & Meier, 2012). Therefore it is necessary to do more exploration about the roles of attention in implicit learning. So far, all previous Artificial Grammar Learning (AGL) studies used visual stimuli. Thus, it remains unclear whether AGL may be due to the presence of a visual regularity. To investigate the generality of effect of selective attention on AGL, we extend the experimental materials to auditory stimuli.
    90 college students were recruited in two experiments. Sequences of letters and sequences of digits with different grammar rules were presented simultaneously through Dichotic listening technology. In Experiment 1, one group of participants were instructed to memorize the sequences of letters; another, the sequences of digits. In Experiment 2, all participants were instructed to memorize the above two types of sequences.
    Results showed that: (1) when only one of the two stimulus dimensions was selected to attend (Experiment 1), participants learned the structure underlying the selected, but not that one of the ignored dimension; (2) when both stimulus dimensions were selected to attend, both structures were learned by participants (Experiment 2). These findings revealed that participants learned only the grammar for the dimensions to which they are attended.

    The results of two studies strongly suggest that AGL occurs with auditory stimuli and visual perceptual learning is not necessary. The effect of selective attention on AGL is applicable across modalities, it is not only suitable for visual stimulus, but also applies to the auditory stimulus.

    Validity and sensitivity analysis of segment recognition task on implicit sequence learning
    YANG Haibo; LIU Dianzhi
    Acta Psychologica Sinica. 2016, 48 (3): 230-237.   DOI: 10.3724/SP.J.1041.2016.00230
    Abstract   PDF (534KB) ( 1750 )
    Sequence generation task, which was developed by Destrebecqz and Cleeremans based on the experimental logics of Process Dissociation Procedure, is one of the prevalent paradigms to disentangle conscious and unconscious knowledge acquisition in the sequence reaction time task. However, previous studies had indicated that the sensitivity of reproduction test is poor than that of recognition test. Moreover, sequence generation leads to variable Response-to-stimulus intervals (RSIs), but constant RSI interferes with the quality representations of the sequence during training phase. In this paper, we tried to development a Segment Recognition Task, which is a 3-element Recognition Test based on the experimental logics of Process Dissociation Procedure and Signal Detection Theory and keeps the RSI constant, in order to better separate conscious and unconscious influences on learning.
    In segment recognition task, the hit means that subjects consciously or unconsciously identify grammatical sequential segment as grammatical, which is approximate to the inclusion test, P (hit)=R+A (1-R).The false alarm means that subjects unconsciously regard ungrammatical sequential segment as grammatical, which is approximate to the exclusion test, P (false alarm) = A (1-R). In the present study, we adopted two experiments to examine the validity and sensitivity of segment recognition task on implicit sequence learning.
    Experiment 1 verified the validity of the segment recognition task, with RSI as an independent variable and sequence generation task as a criterion. The results indicated that the segment recognition task can separate conscious and unconscious influences on learning.

    The purpose of the second experiment was to compare the sensitivity of sequence generation task and segment recognition task on measuring conscious and unconscious knowledge acquisition. We applied the sensitivity analysis of technical economy and included the involvement as independent variables to examine the contributions of consciousness and unconsciousness to implicit sequence learning. The results indicated that the contributions of unconsciousness to high-involved task in segment recognition task were better than that to low-involved task, while the contributions of consciousness to low-involved task segment recognition task were better than that to high-involved task. The results of the present study suggested that the segment recognition task was a valid measurement to capture the knowledge presentation in implicit sequence learning and this task also demonstrated superior sensitivity to sequence generation task.

    The automatic pop-out mechanism of subject performed tasks effect: Evidence from output monitoring
    LI Guangzheng; WANG Lijuan
    Acta Psychologica Sinica. 2016, 48 (3): 238-247.   DOI: 10.3724/SP.J.1041.2016.00238
    Abstract   PDF (428KB) ( 1029 )
    It is commonly established that simple action phrases are retained better when participants were instructed to perform the actions than they learned the phrases by reading. The superior memory performance after motor learning has been called the Subject Performed Tasks effect (SPT effect). Although a great many of researches have tried to account for the SPT effect and four theories had been proposed, none of them was sufficient to explain the enactment effect. However, the processing theories are contradictory with each other. The main contradiction is whether the SPT effect benefited from the automatic retrieval. So the current research applied the “recall-recognition” paradigm to clarify the mentioned divergence.
    We recruited 80 college students to take part in two experiments and used between-subject design to explore the automatic retrieval mechanism by recall - recognition test. The participants were told to memorize the items for later recall by just reading them silently (VTs) or enacting them symbolically after reading (SPTs). An example of an item was given to the participants, which ensured that they were familiar with the task. After presenting the last item, a free recall test was administered which was audio recorded for later examination directly. Then the subjects were given the instructions for the recall-recognition test. Thirty-six words just learned were presented to the subjects on the computer screen. Subjects were instructed to judge whether the words had been recalled or not. The data were analyzed in terms of the serial position of items, by comparing the performance of free recall test and the recall-recognition test.
    With regard to the free recall performance, it displayed that the participants recalled more phrases after finishing the SPTs than the VTs. However, the effects were only appeared in triplets 10, 11 and 12. A primacy and recency effect were observed after learning of a series of items in the VT condition. Whereas, for the SPT condition, no primacy effect but an extended recency effect could be found. In terms of recall-recognition test, the performance was significant worse under the SPTs condition than that under the VTs condition. However, such differences were observed only in blocks of 2~9, 10, 11 and 12. With regard to the category recall performance, there were similar serial position curves for VTs and SPTs, however the performance of recall-recognition was the same with the results of experiment 1.
    The performance of free recall indicated that the SPT effect benefited from the extended recency effect which stems from an automatic pop-out mechanism. However, the recall-recognition test results demonstrated that the retrieval is automatically pop-out under the SPTs condition in blocks of 2~9, 10, 11 and 12. In conclusion, the SPT effects seem to be the result of a significant enhancement of middle and last blocks. The recall-recognition test results were different from the free recall results. It may be that "recall- recognition" paradigm is more sensitive to measure retrieval mechanism of SPT effect.
    Effect of family environment on reading ability: The mediating effects of reading engagement and reading interest among junior high school students
    WEN Hongbo; LIANG Kaili; LIU Xianwei
    Acta Psychologica Sinica. 2016, 48 (3): 248-257.   DOI: 10.3724/SP.J.1041.2016.00248
    Abstract   PDF (545KB) ( 5024 )

     

    Several studies found that family factors were important to students’ reading ability. However, there are few studies discussing the internal mechanism. As students’ individual characteristic, reading interest is one of the mediating factors between family environment and reading ability. Nowadays, investigators have paid more attention to the reading engagement, and some studies show that it is reasonable to believe the mediating effect of reading engagement. Therefore, it’s important to discuss the mediating effect of reading interest and reading engagement on reading ability. The first study we explored was whether reading interest and reading engagement were the single-step multiple mediating variables between family environment and reading ability. The hypothese1: reading interest and reading engagement were single-step multiple mediators between family environment and reading ability. Furthermore, studies show that reading interest had significant influence on reading engagement. The second study was whether reading interest and reading engagement were the multiple-step multiple mediating variables between family environment and reading ability. The hypothese2: the relationship between family environment and reading ability via the influence of reading interest on reading engagement. However, some researchers suggest that there is the interaction between reading interest and reading engagement so the third study was whether the interaction between reading interest and reading engagement could be the mediating effect between family environment and reading ability. The hypothese3: The relationship between family environment and reading ability was in nature through the interaction between reading interest and reading engagement.
    Participants were 574 8th students from two medium counties. In each county, there were 3 common junior high schools selected which were categorized as better, general, poor on the basis of school conditions respectively. And two classes were selected randomly from each school. All students Participated in the reading ability test and reading questionnaire test. After testing, 15 students were excluded from the analyses because of incomplete data in reading ability test and same answers in reading questionnaire test. The final samples included 273 (48.84%) male students and 286 (51.16%) female students. The population was distributed widely. There were 301 (53.85%) students from towns, and 258 (46.15%) students from villages. Participants completed two measures: reading questionnaire (32 items) and reading ability test (24 items). Reading questionnaire included personal information, family environment questionnaire, reading interest questionnaire, reading engagement questionnaire. Reading ability test included four passages which were categorized as narrative (2 passages) and expository (2 passages).
    Preliminary and descriptive analyses showed: The internal consistency for reading questionnaire (family environment, reading interest, reading engagement) was measured by Cronbach’s alpha,they were 0.720, 0.836, 0.792. The internal consistency for reading ability test was.870. In study 1, reading engagement could be a mediator of the relationship between family environment and reading ability. Model fit indices in SEM were χ2 = 575.951, df = 128, χ2/df = 4.500 (p < 0.001), TLI = 0.816, CFI = 0.846, RMSEA = 0.079. In study 2, the relationship between family environment and reading ability was through the effect of reading interest on reading engagement. Model fit indices in SEM were χ2 = 334.698, df = 128, χ2/df = 2.615 (p < 0.001), TLI = 0.915, CFI = 0.929, RMSEA = 0.054. In study 3, The relationship between family environment and reading ability was in nature through the interactive cycle between reading interest and reading engagement. Model fit indices in SEM were χ2 = 334.751, df = 129, χ2/df = 2.595 (p < 0.001), TLI = 0.916, CFI = 0.929, RMSEA = 0.053.

    SEM results are different from previous studies: (1) reading engagement is the mediating variable between family environment and reading interest, reading engagement is the mediating variable between family reading interest and reading performance; (2) The relationship between family environment and reading ability is in nature through the interactive cycle between reading interest and reading engagement. Reading engagement plays a critical role in internal mechanism between family environment and reading ability.

    Patterns of reinforcement sensitivity for non-suicidal self-injury in college students
    YING Mengting; JIANG Guangrong; YU Lixia; LU Ting
    Acta Psychologica Sinica. 2016, 48 (3): 258-270.   DOI: 10.3724/SP.J.1041.2016.00258
    Abstract   PDF (486KB) ( 2985 )
    Reinforcement sensitity, as an important personality trait, affects individual’s emotional, motivational and behavioral process, especially relates to emotion reactivity. Numerous studies have demonstrated that two dimensions of reinforcement sensitivity—BIS sensitivity and BAS sensitivity—exhibit strong associations with emotion management difficulties, which are considered as main characteristics of non-suicidal self-injury. Individuals at the far poles of BIS and BAS dimensions are at increased risk for developing various types of psychopathology, some of which have high correlations with non-suicidal self-injury. In order to investigate the relation between reinforcement sensitivity and non-suicidal self-injury, two studies were conducted.
    Study 1 examined the associations between dimensions of reinforcement sensitivity and non-suicidal self-injury by using self-report measures. The BIS/BAS Scales and the Chinese version of Self-Harm Inventory were used to assess reinforcement sensitivity and non-suicidal self-injury, respectively. Data were collected among 717 college students and were analyzed using ordinal logistic regression. Results showed that BIS/BAS sensitivity significantly predicted non-suicidal self-injury. As to BAS sensitivity, BAS-FunSeeking was the only dimension that contributed to non-suicidal self-injury.
    On the basis of Study 1, Study 2 further explored different patterns of BIS/BAS sensitivity between people who engaged in non-suicidal self-injury for two different purposes. The Q-TASK and the CARROT were used as two behavioral measures of BIS sensitivity and BAS sensitivity, respectively. Two groups of self-injurers (emotion managing vs. novelty seeking) as well as one control group participated in two experiments. Data obtained from 127 participants were analyzed by MANOVA. The Findings indicated that there were different patterns of reinforcement sensitivity for non-suicidal self-injury with different functions. Participants who conducted self-injurious behaviors to manage their own feelings or emotions were characterized by high BIS sensitivity and average BAS sensitivity, whereas participants who engaged in non-suicidal self-injury for seeking novelty were characterized by high BAS sensitivity and average BIS sensitivity.

    In summary, reinforcement sensitivity was a significant predictor of non-suicidal self-injury. It can be hoped that therapeutic interventions for self-injurers based on their patterns of BIS/BAS sensitivity will be designed and proposed before long. More importantly, the different patterns of BIS/BAS sensitivity in self-injurers, which was found in Study 2, indicates different mechanisms behind non-suicidal self-injury. This may shed light on the inconsistent findings regarding non-suicidal self-injury so far. Participants are usually not classified into groups by different functions of their self-injurious behaviors, thus leading to a counteracting effect on potential associations. It is indicated that classification of non-suicidal self-injury is quite necessary and deserves future research attention.

    Perceived societal fairness and goal attainment: The differnet effects of social class and their mechanism
    HU Xiaoyong; GUO Yongyu; LI Jing; YANG Shenlong
    Acta Psychologica Sinica. 2016, 48 (3): 271-289.   DOI: 10.3724/SP.J.1041.2016.00271
    Abstract   PDF (679KB) ( 10918 )
    Goals contribute significantly to the meaning of event and outcome. The attainments of important life goals, such as higher education and employment, are the source of individual well-beings, as well as an effective way to reduce the gap between the rich and the poor. Motivation is one of the most widely studied factors that affect goal attainment. However, according to a series of latest studies, relative to upper-class individuals, lower-class individuals’ goal commitment is more easily affected by societal fairness. Based on the aforementioned studies, the present research intended to explore the effect and mechanism of societal fairness on goal attainment of people from different social classes and hypothesized that goal attainment is affected by (a) the perceived societal fairness, (b) social class, and (c) the interaction between perceived societal fairness and social class. Moreover, we also expected that the interaction effect between perceived societal fairness and social class on goal attainment is mediated by goal commitment.
    We conducted 3 studies to test our hypothesis. The first two studies are the naturalistic design. In Study 1, we measured participants' perceived of social justice, subjective social class, objective social class, and goal commitment and attainment toward educational investment. In Study 2, we randomly assigned participants to two different priming conditions (fairness vs. unfairness), and then measured their subjective social class, objective social class, academic goal commitment, and academic goal attainment. In Study 3, we brought our research question into the lab. This study involves a 2 (fairness: fairness vs. unfairness) × 2(social class: higher-class vs. lower-class) between-subject design. Participants in study 3 completed both fairness and social class manipulation.
    In Study 1, respondents reported higher motivation to invest more educational resource to their child to the extent that they believed that educational rewards were distributed fairly; this effect was stronger for members of lower-class than for members of higher-class. In Study 2, the lower-class participants had higher academic goal commitment and higher academic goal attainment when they were under fair priming condition. However, it did not apply for higher-class participants. In Study 3, the interaction between fairness and social class significantly affected goal commitment, and goal commitment positively affected goal achievement. Moreover, fairness positively related to goal commitment for lower-class members, but not for higher-class members. Taken together, these studies support the hypothesis that social fairness can affect goal attainment through goal commitment and that this connection is significant among participants low in social class but non-significant among participants high in social class.

    The findings of present research suggests that the low class members’ goal commitment and attainment were more susceptible to societal fairness, which is consistent with the ideas of social class psychology who claims that low social class members are situation-oriented and more sensitive to context factors, but those with high social position could freely set their goals and not be limited by the environment. To some degree, our research has provided support for the idea of social class psychology.

    The characteristics of moral emotions of Chinese netizens towards an anthropogenic hazard: A sentiment analysis on Weibo
    YE Yonghao; XU Yan; ZHU Yijie; LIANG Jiongqian; LAN Tian; YU Miao
    Acta Psychologica Sinica. 2016, 48 (3): 290-304.   DOI: 10.3724/SP.J.1041.2016.00290
    Abstract   PDF (837KB) ( 3898 )
    Weibo provides its users a cyber platform to share opinions and show their emotions towards issues at home and abroad. In the process, massive amounts of data are made, and becomes the raw material for sentiment analysis. Previous studies in related fields of computer science and communication focused mainly on developing better sentiment analysis techniques to analyze basic emotions. To add a new perspective, this paper focused on studying the moral emotions expressed toward the “7.23 Wenzhou Train Collision” by Chinese netizens on Weibo. In particular, we analyzed the frequencies of different moral emotions expressed, and related them to the temporal occurrence of different moral events (e.g., statements made by the authority or victims that have moral implications) in the aftermath of the collision, and how different patterns of moral emotions were expressed by different groups including male and female, VIP and non-VIP users.
    First of all, we utilized Weibo API to obtain the Weibo Dataset. Specifically, from July 23rd, 2011 to September 1st, 2011 we used several developer IDs to keep grabbing the public timeline, which is a sample of the real time tweets. Then we used a set of keywords to filter out irrelevant tweets and obtain tweets related to the train accident happened on July 23rd. In total, we got 94,562 valid tweets, among which 21,466 tweets contain users’ information. Secondly, we conducted sentiment classification using K-Nearest Neighbor approach based on the training data labeled by 41 experts. After that, all tweets in the dataset were assigned scores from 0 to 5 for each categories of sentiment and the sentiment evolution chart was drawn. Thirdly, we related the knee points of the chart to the moral events happened during the aftermath of the train collision to identify which emotion was evoked by a certain event. Fourthly, we conducted logistic regression and Robust Maximum Likelihood Estimator (MLR) to analyze the difference of emotional expression among different groups.
    Results indicated that (1) people showed strong moral anger, contempt, disgust, compassion and love during the moral events that happened in the aftermath of the train accident. (2) The pattern of associations between the moral events and the moral emotions supported the Moral Foundation Theory. (3) Men, in comparison to women, were more likely to express anger, disgust and contempt with higher intensity while women exceeded men at both the likelihood and intensity in expressing compassion and love. (4) For compassion and love, organizational VIP users showed higher likelihood as well as intensity than individual VIP users and non-VIP users. But when it came to anger, disgust and contempt, individual VIP users showed higher likelihood and emotional intensity than non-VIP users and organizational VIP users.

    The results are consistent with the Moral Foundation Theory, which postulates that violation of different moral foundation elicits different moral emotions. Moreover, our work is the first one to point out group differences in the expression of moral emotions on Weibo.

    Relationship between moral disengagement in sport and prosocial and antisocial behaviors in Chinese athletes
    WANG Dong; CHEN Zuosong
    Acta Psychologica Sinica. 2016, 48 (3): 305-317.   DOI: 10.3724/SP.J.1041.2016.00305
    Abstract   PDF (560KB) ( 1852 )
    Moral disengagement is a psychological process in which people use different cognitive strategies to justify their potentially problematic actions and to reduce feelings of shame and guilt. Athletes with high level of moral disengagement in sport may show more antisocial behavior and less prosocial behavior than other athletes. To date, most research investigating prosocial and antisocial behavior in sport has focused on behaviors directed at opponents. However, recent studies have also investigated prosocial and antisocial behavior within sport teams. Since the majority of past studies have been conducted in western countries, we know little about moral disengagement of Chinese athletes. Therefore, the purpose of the present study was to develop a valid and reliable measure of sport moral disengagement and to examine the relationship between sport moral disengagement and prosocial and antisocial behavior among Chinese athletes.
    Study 1 sought to develop a sport-specific measure of moral disengagement for Chinese athletes. Forty-eight items were developed based on content analysis of interviews with athletes and mass media reports of transgressions of athletes. Two hundred and fifty-four Chinese athletes were asked to rate their agreement with each item, and their responses were used exploratory factor analyses (EFA). Results showed that 20 items loaded on five factors: (1) conduct reconstrual, (2) advantageous comparison, (3) euphemistic labeling, (4) dehumanization, and (5) non-responsibility. After this, another 283 Chinese athletes filled in the questionnaire. The confirmatory factor analysis (CFA) indicated that the 5-factor solution was stable and the internal reliability of each of the factors was acceptable.
    Study 2 tested the relationship between moral disengagement in sport and prosocial and antisocial behavior towards teammates and opponents. Three hundred and six Chinese athletes responded to the 20-items moral disengagement questionnaire developed in Study 1 and the Chinese version of prosocial and antisocial behavior in sports scale (PABSS). Results from multivariate analysis of variance (MANOVA) and hierarchical regression analysis (HRA) showed that gender, type of sport, and competition level can predict antisocial behavior towards teammates and opponent (R2=12.88, p < 0.01; R2 =41.31, p < 0.01) better than prosocial behavior towards teammates and opponents (R2 =5.37, p < 0.01; R2 =6.23, p < 0.01). Type of sport was the strongest predictor which negatively predicted antisocial behavior towards teammates and opponents (β = −0.29, t = −5.32, p < 0.01; β = −0.49, t = −9.94, p < 0.01). After controlling for gender, type of sport, and competition level, non-responsibility was the strongest predictor which can significantly and negatively predict prosocial behavior towards teammates and opponents (β = −0.31, t = −5.81, p < 0.01; β = −0.28, t = −5.27, p < 0.01); euphemistic labeling was the strongest predictor to positively predict antisocial behavior towards teammates (β = 0.23, t = 4.16, p < 0.01); conduct reconstrual was the strongest predictor which could significantly and positively predict antisocial behavior towards opponents (β = 0.25, t = 5.33, p < 0.01).= 0.06, F(3, 302)= 0.05, F(3, 302)= 0.29, F(3, 302) = 0.11, F(3, 302)

    The present study contributes to research into moral disengagement in sports by developing a valid measurement instrument for moral disengagement in Chinese athletes, and by analyzing the relationship between sport moral disengagement and prosocial and antisocial behavior towards teammates and opponents in Chinese athletes.

    Factors affecting the classification accuracy of reparametrized diagnostic classification models for expert-defined polytomous attributes
    ZHAN Peida; BIAN Yufang; WANG Lijun
    Acta Psychologica Sinica. 2016, 48 (3): 318-330.   DOI: 10.3724/SP.J.1041.2016.00318
    Abstract   PDF (732KB) ( 976 )
    Diagnostic classification assessment (DCA) utilizes latent class models to provide fine-grained information about students’ strengths and weaknesses in the learning process. In the past decades, extensive research has been conducted in the area of DCA and many statistical models based on a probabilistic approach have been proposed. At present, several diagnostic classification models (DCMs) for dichotomous attributes exist, which include the deterministic inputs, noisy “and” gate (DINA; Junker & Sijtsma, 2001); the deterministic inputs, noisy “or” gate (DINO; Templin & Henson, 2006); and the linear logistic model (LLM; Maris, 1999). In contrast, only a few DCMs can be used to deal with the polytomous attributes, such as the model based on the ordered-category attribute coding (OCAC; Karelitz, 2004), and the polytomous generalized DINA (pG-DINA; Chen & de la Torre, 2013).
    Polytomous attributes, particularly those defined as part of the test development process, can provide additional diagnostic information. The present research proposes three reparametrized reduced models of pG-DINA (Chen & de la Torre, 2013), which include the reparametrized polytomous attributes DINA (RPa-DINA), the reparametrized polytomous attributes DINO (RPa-DINO), and the reparametrized polytomous attributes LLM (RPa-LLM). Furthermore, to better understand the classification accuracy of the new models, the impact of 6 factors was investigated, namely, the number of polytomous attributes, the highest level of polytomous attributes, the correlations among polytomous attributes, the hierarchical structure, the sample size, and the number of items. Results of the simulation study indicated that:
    (1) more polytomous attributes led to lower classification. Their effects, in descending order, were the RPa-LLM, the RPa-DINO, and the RPa-DINA. Less than 5 polytomous attributes used in empirical research is suggested;
    (2) for the number of attribute levels, more levels resulted in worse performance. Less than 4 levels within one attribute used in empirical research is suggested;
    (3) the higher the correlations among polytomous attributes, the higher the classification accuracy would be;
    (4) different hierarchical structure had different influences on the classification accuracy. No matter what structure we had, the performance of RPa-DINA was quite well behaved. However, other 2 models, especially the RPa-DINO, were recommended for the analysis of response data from independent hierarchical structure;
    (5) the sample size has little impact on the classification accuracy; and

    (6) the number of items was inversely proportional to the classification accuracy.

Copyright © Acta Psychologica Sinica
Support by Beijing Magtech