ISSN 0439-755X
CN 11-1911/B


    30 July 2010, Volume 42 Issue 07 Previous Issue    Next Issue

    For Selected: Toggle Thumbnails
    The Influence of Working Memory Span on the Accentuation Effects in Discourse Comprehension
    ZHANG Xuan,YANG Yu-Fang
    2010, 42 (07):  727-734. 
    Abstract ( 1563 )   PDF (339KB) ( 2675 )  
    Working memory capacity is of central importance for many complex cognitive tasks such as language comprehension. Previous research has focused on how working memory capacity affects syntactic, semantic, and pragmatic processing. It is not clear whether working memory capacity also affects prosody. At the prosodic level, the correspondences between accentuation and information structure have been proven to exert an influence on discourse comprehension. This study explored the influence of working memory capacity on accentuation effects in discourse comprehension from an individual difference approach.
    High (n=27) and low (n=23) working memory span participants were selected by Reading Span Test from 106 college students. Sentence-by-Sentence Auditory Moving Window Paradigm was employed to measure the effects of accentuation on discourse comprehension, with 24 discourses (each containing 5 sentences) as materials. The on-line processing time of discourse comprehension, measured by difference time, was compared among 3 conditions: consistent accentuation condition, inconsistent accentuation condition and controlled condition.
    Low working memory span participants showed speeded on-ling processing in the consistent condition, and slowed processing in the inconsistent condition. High span participants, on the other hand, showed neither effect. High and low span participants did not differ significantly in either the consistent or neutral accentuation conditions. In the inconsistent accentuation condition, however, low span participants spent significantly more time on the on-line processing of spoken discourse than high span participants.
    These results suggest that:
    (1) Accentuation effects in discourse comprehension are influenced by working memory capacity, and the influence can be explained by the controlled attention view of working memory. 
    (2) Practical implications of this study would be to call attention to the appropriate employment of prosody such as accentuation, especially during the verbal communication with people who have lower working memory capacity.
    Related Articles | Metrics
    Differences of Emotional Words in Implicit and Explicit Memory Tests: An ERP Study
    LI Yue -Ting,LI Qi,GUO Chun-Yan
    2010, 42 (07):  735-742. 
    Abstract ( 1818 )   PDF (444KB) ( 2768 )  
    The previous studies have shown that for negative and positive stimuli, the deep physiological mechanisms are different in the processing of unconscious encoding and retrieval. Nonetheless, the responses to the emotional words in the retrieval phase have shown contradictions in the conscious explicit memory tests. Due to the differences of the encoding tasks and experimental paradigms, however, we cannot attribute the performance differences to stimulus type differences. Therefore, the same task was conducted in the encoding phase to investigate how different emotional stimuli impact the performance in implicit and explicit memory tests. Furthermore, many researchers have proved that the encoding of the emotional words is implicit, and the implicit memory and explicit memory is dissociated, we also want to explore whether the processing of emotional words will influence the implicit or explicit retrieval.
    An ERP study was conducted to explore the differences between positive words and negative words in implicit and explicit memory tests, adopting the study-test paradigm. 16 right-handed participants were employed in the study. There were three phases in each block: (1) The study phase required the participants to decide whether a word was a noun or an adjective with different key-pressing response. After the study phase, the participants were asked to substract 3 from a 3-digit number shown on the screen.(2) The lexical decision was conducted after the study phase in which participants need to decide the item was a word or a pseudo-word by pressing the keys as soon as possible (3) The recognition test asked the participants to judge whether they had saw the presented word in the study phase and make a key-pressing response. Each stimulus lasted 800 ms, ISI 1100 ± 100 ms.
    The results showed that negative words were associated with a more positive ERP than positive words in the lexical decision. There was an ERP dissociation of retrieval formats between the implicit and the explicit memory tests during 450~900 ms, that is, there was a significant difference between negative words and positive words in the implicit memory whereas there was no difference between them in the explicit memory.
    So we can come to the conclusion that the implicit retrieval was sensitive to emotional stimuli while the explicit retrieval was not. The strong attention-grabbing of negative words was shown in the implicit retrieval phase.
    Related Articles | Metrics
    Mechanism of Delayed-JOL Effect
    CHEN Gong-Xiang,ZHANG Cheng-Fen,SU Ya-Wen
    2010, 42 (07):  743-753. 
    Abstract ( 1526 )   PDF (379KB) ( 1951 )  
    There are two kinds of hypotheses regarding to delayed judgment of learning (JOL) effect, which are memory hypothesis and metamemory hypothesis. Two experiments were conducted to investigate the controversy between memory hypothesis and metamemory hypothesis, and the mechanism of monitoring-dual-memories hypothesis (MDM, Nelson & Dunlosky, 1991), respectively.
    Both the memory hypothesis and the metamemory hypothesis acknowledge that the accuracy of JOL increases when in delayed-JOL condition than in immediate-JOL condition. However, the explanation of these two hypotheses on the mechanism of this phenomenon differs substantially. The metamemory hypothesis believes that delayed-JOL effect is duo to improvement in metamemory, while the memory hypothesis supposes that delayed-JOL simply affects memory, not metamemory. We believed that these two hypotheses are not either-or. Namely, delayed-JOL affects memory as well as metamemory.
    In the first experiment, we removed the memorial difference between delayed-JOL and immediate-JOL by setting in a preliminary test between JOL and final test. The results showed that the memory scores in immediate-JOL condition and delayed-JOL condition had no significant difference. However, the accuracy of delayed-JOL was substantially high than the accuracy of immediate-JOL in a preliminary test. This result confirmed that the delayed-JOL indeed improved the accuracy of metamemory, it affected memory as well as metamemory.
    Now that we have proved that the delayed-JOL improved metamemory in the first experiment, we continued to study the mechanism of this improvement in the second experiment. There are three types of metamemory hypotheses, MDM hypothesis is most influential and controversial. There are many researches supported this hypothesis, but Kelemen and Weaver′s (1997) study challenged it. However, they can only determine the removement of short-term memory (STM) through theoretical speculation in traditional experimental procedure, whether the STM was removed or not and the extent of removement could not be measured. We could observe the score of pre-judgment recall by using PRAM method. The results of experiment showed that the distractive assignment could not remove STM effectively. We found that the accuracy of JOL was related to the removement of STM through correlation analysis. The accuracy improved with the extent of removement of STM. So we believed that the delayed-JOL effect was duo to removement of STM thoroughly. Our study confirmed as well as extended the MDM hypothesis.
    Related Articles | Metrics
    3 to 5 Years Old Children’s Behavioral and Verbal Performances in Level 1 Perspective-Taking
    ZHAO Jing,WANG Lu,SU Yan-Jie,CHAN Raymond C. K.
    2010, 42 (07):  754-767. 
    Abstract ( 1624 )   PDF (255KB) ( 2173 )  
    The study aimed to investigate the level 1 perspective-taking of 3 to 5-year-old children and to explore the potential differences between their behavioral and verbal performances. Specifically, Experiment 1 adopted the “one-way mirror” paradigm to capture the performances of 54 children. Logistic regression analysis showed that children’s behavioral performances were almost the same across ages (B = -0.225, Wald = 0.38, p = 0.541). ANOVA analysis also found that there was no discrepancy between behavioral and verbatim data. Given the limitations of the scoring system and lacking control of natural preference in the behavioral paradigm, we improved it in Experiment 2 by applying a hide-and-seek game to another group of 60 children. Three-way ANOVA analysis indicated significant interaction among task type, age and gender (F (2, 48) = 3.55, p = 0.037, η2 = 0.13). These results suggested that the behavioral performances improved from 3 to 5-year-olds in both girls and boys, with boy s’ verbal responses improved to a lesser extent than girls’. Taken together, these findings suggested that a behavioral task might provide additional information to tasks designed merely to capture verbal responses when assessing children’s level 1 perspective-taking.
    Related Articles | Metrics
    Temperament and Early Teacher-Child Relationships: The Moderating Role of Family Affective Environment
    ZHANG Xiao,WANG Xiao-Yan,CHEN Hui-Chang
    2010, 42 (07):  768-778. 
    Abstract ( 1937 )   PDF (540KB) ( 3505 )  
    Temperament has been regarded as crucial in child development. Two models dominate current research on the role of early temperament in children’s cognitive, social, and emotional development, namely, direct effect model and interactive effect model. Although research has placed great emphasis on the impact that temperament has on peer competence, relatively little attention has been paid on the role of temperament in children’s competence to establish positive relationships with teachers. The aim of the current study was to investigate the role of early temperament in young children’s relationships with teachers. More importantly, both direct effect and interactive effect models were tested in order to fully understand the role of temperament in children’s teacher-child relationships.
    Based on a two-year and two-wave longitudinal sample of 118 preschoolers and their parents and teachers, the current study examined the relations of temperamental activity level, rhythmicity, approach/withdrawal and family affective environment to teacher-child relationships by using Parent Temperament Questionnaire (PTQ), Family Environment Scale (FES), and Student-Teacher Relationship Scale (STRS). Hierarchical regression analyses were conducted to investigate the unique effects of temperament and the interactive effects of temperament and family affective environment on teacher-child relationships.
    Results showed that approach/withdrawal, rhythmicity, and activity level were associated significantly with teacher-child relationships two years after children’s preschool entrance, and that these temperamental dimensions accounted for 3~13% of the variances in these teacher-child relationships. Moreover, the temperamental dimensions also interacted with family affective environment to influence teacher-child relationships at both waves, which accounted for 4~10% of the variances in the relationships.
    These findings highlight the importance of examining the interactive effect of temperament and environmental factors on children’s teacher competence, and expand the existing research on temperament and teacher competence by testing both direct effect and interactive effect models.
    Related Articles | Metrics
    Attention Biases of Undergraduate Women with Fat Negative Physical Self: Orienting or Maintenance
    FENG Wen-Feng,LUO Wen-Bo,LIAO Yu,CHEN Hong,LUO Yue-Jia
    2010, 42 (07):  779-790. 
    Abstract ( 1679 )   PDF (418KB) ( 3077 )  
    Fat Negative Physical Self (FNPS) is defined as negative cognition, emotion and behavior regarding the self as fat. There are some studies on attention biases for fat-related information among FNPS individuals. However, researchers assume that the central feature of attention biases in the FNPS is attention orienting, which means that fat-related information attracts the initial attention of FNPS individuals. The present study raises doubt about this view. Another possibly more salient attention bias in FNPS individuals is of that attention maintenance, wherein fat stimuli are difficult to disengage from after they have been detected.
    Three experiments were conducted to test the hypothesis. Participants were assigned to experimental and control groups from their scores on the Negative Physical Self Scale-Fat (NPSS-F, Chen, Jackson, Huang, 2006). Experiment 1a used a cuing paradigm and the SOA was set to 300ms, in order to study orienting versus disengagement components of attention bias. Experiment 1b used the same paradigm as experiment 1a, with the exception that the cueing materials were body figure pictures and vase silhouette. In experiment 2, we modified the paradigm so that participants had to identify letters that might be located around pictures which were presented on the center of the monitor.
    The results of experiment 1a and 1b were consistent with the hypothesis. The results of Experiment 2 suggested that attention was maintained on the fat body figure longer than on the thin body figure. However, another unexpected result was that time focused on the normal body figure was also longer than that of the thin body figure. Further analyses indicted results of Experiment 2 were not a function of BMI but instead a result of heighted fatness concern.
    In conclusion, we found that attention of female college students with high FNPS levels reflected more maintenance than initial orienting towards fatness stimuli.
    Related Articles | Metrics
    The Developing of Faking Detection Scale in Application Situation
    LUO Fang,LIU Hong-Yun,ZHANG Yue
    2010, 42 (07):  791-801. 
    Abstract ( 1150 )   PDF (599KB) ( 2210 )  
    Using personality tests in personnel selection becomes increasingly popular in industries in China. However, the validity of personality tests is challenged due to the possibility of getting faking answers (Morgeson et al, 2007). Methods dealing with faking in personality tests are available. The most commonly used one is to detect faking by using social desirability (SD) scale, and then calibrate the results to remove the faking effect.
    However, authors of this study argue that faking is conceptually different from socially desirable responding (SDR), and using SD Scale to measure faking effect is inappropriate. The reasons include: (1) although both faking and SDR are the tendency to exaggerate, the former is influenced by job desirability, while the later is influenced by social desirability. (2) SD Scale can not measure Impression Management and Self-Deception independently and has poor construct validity (Paulhus, 2002).
    This research developed a “Faking Detection Scale for General Positions” in seven steps based on the procedures used to develop the SD scale and the special tributes of faking.
    First, behavior descriptions which are strong job-desirable and extremely infrequent or high frequency were collected and rewritten into 114 items. Second, trained raters rated the desirability and occur-probability of these items and selected 50 items for the draft scale. Third, an experimental study was conducted with 679 subjects to further select items. Subjects were paid for 20 RMB if they successfully faked in simulated selection situations. As a result, 20 items which sufficiently measure faking and are not contaminated by Self-Deception remained in the scale. Forth, principal component analysis was used to examine the construct validity of this scale. Results showed the first factor explained 55% of the total variance. Fifth, multivariate generalization theory was used to study the optimal number of items included in the scale. The results showed that when 10 items remained in the Faking Detection Scale, the scale has good reliability, with G=0.823 and φ=0.818. Sixth, the scale’s validity was further examined using logistic regression. The Cox&Snell R Square and Nagelkerke R Square were .386 and .468, repectively. These results showed that the scale was valid and can effectively detect faking candidates. However, there is a dilemma when setting the cut-off score, because it is impossible to decrease the false judgment and false selection rate at the same time. Results from this study showed that setting up higher cut-off score to detect serious faking candidates and exclude unqualified candidates using select-out mode can help organization make valid recruitment decisions and accomplish personnel selection goals. Finally, this scale’s validity was verified in real occupational selection situations with 234 subjects. Faking Detection Scale (effect size was 1.163) is far more sensitive than personality scale (the largest effect size was .767) and can more effectively measure faking effect. This scale has been validated and can effectively detect faking candidates in real occupational selection situations.
    By exploring the quality of items, it is found that most faking detection items described common behaviors in daily life and repeatedly emphasized in work-settings. Subjects showed a high consciously inflation on these items when they were faking; when faking did not happen, they would unconsciously exaggerate the reaction to a less extent because of Self-Deception. So the scale could sufficiently measure faking effect with less contaminative by Self-Deception.
    As a conclusion, this study differentiated faking from SDR and developed a scale to detect faking. This scale has 20 items which is very sensitive to faking. Its validity was verified in simulative and one real occupational selection situation. More studies are needed to further explore whether this scale can be applied to varied situations.
    Related Articles | Metrics
    The Impact on Ability Estimates of Predicted Parameters from Cognitively Designed Items in a Computerized Adaptive Testing Environment
    YANG Xiang-Dong
    2010, 42 (07):  802-812. 
    Abstract ( 925 )   PDF (537KB) ( 1260 )  
    Automatic item generation has become a promising area in recent studies. In automatic item generation, items with targeted psychometric properties are generated during testing. The feasibility of automatic item generation lies in the fact that items are generated from a set of observable item stimulus features, which are mapped onto the cognitive variables underlying the item solution and are calibrated through cognitive psychometric models. Parameters of a generated item can then be predicted from the specific combination of the calibrated item stimulus features in the item. Predicted item parameters, compared to those calibrated from empirical data, involve more complex sources of uncertainty. Although the relationship between sufficiency of the cognitive model of item solving and the adequacy of item parameter prediction can be theoretically justified, the degree to which such predicted parameters impact various aspects of testing, however, is an empirical question and needs to be explored.
    This paper investigated the impact of predicted item parameters on ability estimates in a computerized adaptive testing environment, based on abstract reasoning test (ART) items which were generated using cognitive design system approach (Embretson, 1998). The item bank contained 150 items with two sets of item difficulties, of which one is predicted from the item design features and the other is calibrated from sample data. Each of the 263 subjects participated in the study received two subtests, of which one was based on predicted parameters, the other calibrated parameters. The item bank was split into two parallel halves based on predicted item parameters to prevent items in the bank from repeat administrations within subjects. Subjects were randomly assigned into one of the four testing procedures resulting from the combinations between parameter types (predicted versus calibrated) and item bank (first half versus second half).
    Results of the study showed a clear regression-to-the-mean effect of the predicted item parameters, compare to its counterpart of calibrated item parameters. Inward biases of ability estimates from subtest using predicted item parameters were observed when ability estimates were compared across different subtests within subjects. Compared to its counterpart using calibrated parameters, standard errors of ability estimates were larger for those from the subtest using predicted item parameters in the mid-range of the scale, where regression-to-the-mean effect of the predicted item parameters is minimal, and were smaller in the rest of the scale, possibly due to joint impact of increased uncertainty of predicted item parameters, estimation biases and limitation of the item bank at various levels of ability scale. When ability was estimated from the same subtest using different types of item parameters, very high correlation (.995) were obtained and no biases were observed throughout almost the entire scale. Standard errors of ability estimates were larger for predicted parameters yet the differences were small.
    Related Articles | Metrics
    Parameter Estimation of the Raw Item in Computerized Adaptive Testing
    YOU Xiao-Feng,DING Shu-Liang,LIU Hong-Yun
    2010, 42 (07):  813-820. 
    Abstract ( 1300 )   PDF (410KB) ( 1328 )  
    Along with the development of computer technology and increasing needs of individual learning, Computerized Adaptive Testing (CAT) has received more and more attention. However, the problem of the test security is becoming a new challenge to the CAT, for example, excessive exposure of the test items might weaken the efficiency and equity of the CAT. Therefore, constructing a large-scale and high-quality CAT item bank is highly demanded
    The traditional method of constructing a CAT item bank involves four steps: first, the items are developed by the experts according to the test specification and blueprint; second, a representative sample of examinees is recruited to take the test; Then, the calibration of item parameters and testing of the model-data fit will be conducted afterward. Finally, items of high-quality and its item parameters will be added into the item bank based on the pre-analysis.
    However, several problems exist in the above traditional method. First, a large number of test takers are needed, which is time-consuming and expensive. Second, the security of the items cannot be guaranteed in the pilot test. Third, complicated techniques of equating may have to be used to construct a good item bank, which may also affect the security of the items, especially when the anchor items are repeatedly used in the test equating process.
    Thus, if the raw items can be seeded in the CAT process and the item parameters combined with the examinee abilities estimated at the same time, it will be significant for the construction of CAT item bank. The research in this area has not been widely conducted and reported in the domestic journals, although this issue has been a big topic for foreign researchers. Thus, This study aims to explore how to insert raw items and estimate the item parameters in CAT and investigate the efficient strategies with high security for enlarging the item bank. A new online calibration method is proposed based on the principle of adaption in CAT, provided there is a small-scale item pool. And the formula of initial value during the stage of the iteration is determined. When the ability parameter has been estimated, the conditional maximum likelihood estimation(CMLE) will be implemented to estimate item parameters in the raw items. First, the difficulty parameter could be obtained through CMLE. Second, the discrimination parameter could be estimated with the abilities and difficulty parameter and the initial value of the discrimination parameter could be gained through CMLE. Third, the difficulty parameter could be estimated when the discrimination parameter and abilities are known .Repeat the second and the third steps until the stop condition is satisfied.
    Simulation of Monte Carlo has been employed to estimate the parameters of raw items with the one-parameter Logistic and two-parameter Logistic models in the study and good result has been gained.
    Related Articles | Metrics