因子分析元分析的三阶段整合模型

因子分析元分析(meta-analysis of factor analyses, MAFA)整合了因子分析和元分析两种方法学传统。开展MAFA方法学研究可结束过去几十年MAFA研究缺乏根据的状况, 服务心理学等领域内的知识生产。三阶段整合模型认为MAFA包括数据准备、数据合成和数据分析三个主要阶段。模型及数据合成技术的有效性将用心理测量学方法, 包括蒙特卡洛模拟研究进行检验。研究结果将阐明正确运用MAFA的一般规则、操作步骤和注意事项, 进而增进我们对MAFA的认识, 丰富应用心理等领域的方法学体系, 为MAFA方法的推广应用奠定基础。     

Psychometric Properties of the Locus of Control of Behaviour Scale (LCBS) among Australians Seeking Alcohol and Other Drug (AOD) Treatment

Despite a paucity of studies evaluating the psychometric properties of the Locus of Control of Behaviour Scale (LCBS), it continues to be widely used in behavioural research. The present study sought to redress this gap in the literature. The 17‐item LCBS was administered to 373 Australians attending Alcohol and Other Drug (AOD) treatment agencies in the northern metropolitan region of Perth. Confirmatory factor analyses were conducted in order to determine which of several plausible measurement models provided the best fit for the data. A unidimensional model, recommended by the authors of the LCBS, and a two‐dimensional (Internal Locus of Control vs External Locus of Control) model provided poor fits. Other multidimensional models, differing only in the dimensionality of the externality factor, were also tested. A multidimensional model consisting of an Internal Locus of Control factor and four component External Locus of Control factors provided the best fit; however, the fit is probably best described as “reasonable” rather than “good.” A subsequent exploratory factor analysis using parallel analysis indicated a cohesive internality factor; however, the externality factor showed a tendency to fragment into smaller components. Results were discussed in terms of the problematic externality factor.     

Relational Spirituality and Dealing With Transgressions: Development of the Relational Engagement of the Sacred for a Transgression (REST) Scale

In the study of spirituality and forgiveness, researchers have begun to look at how dynamic spiritual experiences influence forgiveness. In three studies, we develop the Relational Engagement of the Sacred for a Transgression (REST) Scale, which assesses the extent to which victims actively engage a relationship with the Sacred to deal with a specific transgression. In Study 1, exploratory factor analysis was used to determine the factor structure of the REST. In Study 2, the factor structure was replicated using confirmatory factory analysis. The REST was correlated with religious commitment and negatively related to avoidant attachment to God. In Study 3, evidence supporting the scale's construct validity was adduced. REST scores were correlated with other appraisals of relational spirituality. Structural equation modeling was used to compare theoretical models. REST scores were related to dedication to the Sacred and viewing the transgression as a desecration. In addition, REST scores were positively related to empathy, which was positively related to forgiveness. Furthermore, REST scores predicted forgiveness above and beyond other appraisals of relational spirituality.     

Further evidence of the heterogeneous nature of impulsivity

‘Impulsivity’ refers to a range of behaviours including preference for immediate reward (temporal-impulsivity) and the tendency to make premature decisions (reflection-impulsivity) and responses (motor-impulsivity). The current study aimed to examine how different behavioural and self-report measurements of impulsivity can be categorised into distinct subtypes.Exploratory factor analysis using full information maximum likelihood was conducted on 10 behavioural and 1 self-report measure of impulsivity.Four factors of impulsivity were indicated, with Factor 1 having a high loading of the Stop Signal Task, which measures motor-impulsivity, factor 2 representing reflection-impulsivity with loadings of the Information Sampling Task and Matching Familiar Figures Task, factor 3 representing the Immediate Memory Task, and finally factor 4 which represents the Delay Discounting Questionnaire and The Monetary Choice Questionnaire, measurements of temporal-impulsivity.These findings indicated that impulsivity is not a unitary construct, and instead represents a series of independent subtypes. There was evidence of a distinct reflection-impulsivity factor, providing the first factor analysis support for this subtype. There was also support for additional factors of motor- and temporal-impulsivity. The present findings indicated that a number of currently accepted tasks cannot be considered as indexing motor- and temporal-impulsivity suggesting that additional characterisations of impulsivity may be required.     

An examination of the factorial dimensionality of the Marlowe Crowne Social Desirability Scale

Using a sample of 555 undergraduate students, the present study examined the factorial dimensionality of the Marlowe-Crowne Social Desirability Scale (MCSDS). In particular, using confirmatory factor analysis, the study explored the goodness of fit of both one and two factor models, the latter of which were based upon the work of Paulhus and Reid (1989). Findings indicate that while both one and two factor models demonstrate many features of adequate fit, the two factor model emerged as superior. The study concludes with a discussion of the meaning of the results and suggestions for future research.     

Validation of the Adolescent Concerns Measure (ACM): evidence from exploratory and confirmatory factor analysis

This article reports the development and initial validation of scores obtained from the Adolescent Concerns Measure (ACM), a scale which assesses concerns of Asian adolescent students. In Study 1, findings from exploratory factor analysis using 619 adolescents suggested a 24-item scale with four correlated factors--Family Concerns (9 items), Peer Concerns (5 items), Personal Concerns (6 items), and School Concerns (4 items). Initial estimates of convergent validity for ACM scores were also reported. The four-factor structure of ACM scores derived from Study 1 was confirmed via confirmatory factor analysis in Study 2 using a two-fold cross-validation procedure with a separate sample of 811 adolescents. Support was found for both the multidimensional and hierarchical models of adolescent concerns using the ACM. Internal consistency and test-retest reliability estimates were adequate for research purposes. ACM scores show promise as a reliable and potentially valid measure of Asian adolescents' concerns.     

Factor structure of the FACES-III in Japanese university students

We performed a confirmatory factor analysis for the Family Adaptability and Cohesion Evaluation Scale (FACES-III) using 3865 university students in Japan. We confirmed two factors as in the original version, but there were slight differences with the original one. Some items were deleted in order to obtain sufficient goodness-of-fit indexes in a series of confirmatory factor analyses. Moreover, items 11 "Our family, changes its way of handling tasks," 20 "Parents and children discuss in our family," and 15 "We shift household responsibilities," which were originally categorized as "adaptability" items, were loaded on the "cohesion" factor.     

Examining the multidimensionality of approaches to learning through the development of a revised version of the Learning Process Questionnaire

Aim: This study aimed to produce a revised two‐factor version of the Learning Process Questionnaire (R‐LPQ‐2F) with deep and surface approach scales, measured by a reasonably small number of items, suitable for use by teachers in secondary schools to evaluate the learning approaches of their students. Method: A set of 41 items was derived, with modification, from the original version of the LPQ and from items used to develop the revised version of the SPQ. These items were tested using reliability procedures and confirmatory factor analysis and items were deleted until scales were of a suitable length and confirmatory factor analysis indicated a good fit to the intended two‐factor structure. Sample: The sample consisted of 841 students from 20 secondary schools in Hong Kong. Results: The final two‐factor version of the questionnaire had good Cronbach alpha values and reasonable goodness‐of‐fit values for the confirmatory analysis. There was a much better fit, though, to a hierarchical structure with motive and strategy subscales for each approach, each of which, in turn, had two subcomponents. Conclusion: Approaches to learning have a hierarchical dimensionality with motive and strategy elements. Each motive and strategy element is itself multidimensional. The results are used to question the conventional approach to the testing and acceptance of instruments, which place sole reliance upon reliability tests. The use of confirmatory factor analysis is recommended as a routine procedure in the development and testing of instruments.     

Stability over age and sex of children's responses to embedded figures test

240 children in Grades K to 4 were administered the entire Children's Embedded Figures Test to determine whether similar factor structures are obtained from samples of different sex and age as well as when the population is randomly split. Data were analyzed by principal axis factor analysis and relevant factor structures were compared using Veldman's 1967 program "Relate". There was one stable factor which was consistent across sex and accounted for about 19% of the variance. Results were inconsistent across grades and indicated a possible lack of agreement of factor structures.     

Factor analysis to examine psychometric properties of family functioning measures

To assess the relationship of family functioning to problem behaviors and alcohol and drug use among youth, researchers must test the effects of interventions using suitably constructed and psychometrically sound scales. This study evaluated whether originally calculated coefficients alpha underestimate the reliability of the family functioning measures given. Through exploratory factor analysis, estimates of alternative internal consistency reliability which might improve the estimate of reliability were examined. Responses of 755 adults from Strengthening Multi-ethnic Families and Communities were analyzed. Coefficients alpha for two scales were modest (alpha = .68 and alpha = .75), and factor analysis indicated that the scales were multidimensional. After exploratory factor analysis, the reassessment of reliability based on the extracted factors indicated an overall increase in the coefficients alpha.     

A further examination of the constructs underlying assessment center ratings

Previous factor analytic work (Sackett and Dreher, 1982) has found exercise factors, rather than the expected dimension factors, underlying ratings made in managerial assessment centers. The present paper responds to criticisms that inappropriate analyses were conducted and that unrepresentative assessment centers were studied. Exercise factors were found to predominate in two new sets of assessment data which we obtained and in three studies published by other authors, thus indicating that the results were not idiosyncratic to the three centers studied by Sackett and Dreher. Analyses using confirmatory, rather than exploratory, factor analysis also support the dominance of exercise factors. Although dimension factors were found in three of four data sets studied, variables loading on a dimension factor consistently loaded more highly on an exercise factor as well.     

Validation of the acculturative stress inventory for children

The psychometric properties of the Acculturative Stress Inventory for Children (ASIC) were examined with a sample of 139 Hispanic children. This self-report instrument was studied using the items of the Societal, Attitudinal, Familial, and Environmental Acculturative Stress Scale for Children (SAFE-C; Chavez, Moran, Reid, & Lopez, 1997), which had face validity for acculturative stress. The psychometric properties of the ASIC were examined by conducting an exploratory principal-axis factor analysis, which yielded a two-factor solution. The first factor includes 8 items assessing perceived discrimination and the 4 items of the second factor measure immigration-related experiences. Evidence of internal consistency and test-retest reliability for the 2 factors and the total scale were found to be adequate. In addition, sufficient estimates of convergent, discriminant, and predictive validity were obtained. Implications for future research of acculturative stress in children involving the ASIC are discussed.     

Exploratory factor analysis of the Human Behavior Rating Scale: a rural population

The construct validity of Eaves' Human Behavior Rating Scale (HBRS) was investigated. The HBRS is a scale with 91 Likert-type items designed to measure five factors: arousal (persistence and curiosity), affect (externalizing and internalizing), and cognition. Forty-four teachers of Grades 6, 7, and 8 from two low socioeconomic and rural southern counties completed the HBRS for 320 of their students. Three parcels were submitted for analysis for each of the five dimensions. A five-factor, then a two-factor, solution were analyzed using exploratory factor analysis. A principal-axis factor analysis was employed and an oblique promax rotation was applied. The results supported the five-factor solution with 90.2% of the total variance accounted for when compared to the two-factor solution which recovered 73.4% of the total variance. Intercorrelations between factors ranged from .15 to .76 for the five-factor solution.     

Measuring positive,psychological rewards: The validation of the Intrinsic Work Rewards Scale

This study sought to validate the Intrinsic Work Rewards Scale (IWRS) using a transnational sample. Respondents were 486 non-profit employees from Australia, South Africa and the United States of America (females = 72.0%; managerial/ supervisory job level employees = 57.4%). Data analysis included reliability testing, exploratory and confirmatory factor analysis, and concurrent validity testing with established measures to which the IWRS is theoretically linked. The scores from the IWRS obtained an overall acceptable reliability coefficient of 0.86. Factor analysis confirmed its five factor structure, and correlations indicated that intrinsic rewards are positively related to work engagement, and negatively related to intention to quit. The IWRS appears to yield reliable scores for human resource managers advising in the non-profit employment sector.     

Paper-and-pencil and web-based testing: the measurement invariance of the Big Five personality tests in applied settings

This study investigates the measurement equivalence of a five-factor measure of personality across two groups applying for jobs, who completed the same questionnaire using either a paper-and-pencil (n = 429) or a web online answer format (n = 651). The data were collected using the Big Five Questionnaire-2 (BFQ-2; which is a measure of the Five Factor Model) of personality traits. Multiple-group confirmatory factor analysis was used to test for the equivalence of factor covariance and mean structures of the BFQ-2. Findings suggested that the Five Factor Model scales have the same measurement unit and origin across applicants using different administration modes. However, latent means were slightly higher for applicants who responded in a web and unproctored condition than for applicants who completed a paper-and-pencil version of the same test. Practical implications for personality assessment are discussed.     

An empirical investigation of a theoretical model for mathematical creativity

In this exploratory study, a theoretical model proposed by Sriraman (2005) consisting of five theoretical principles for optimizing creativity in a K–12 setting was investigated empirically. This was accomplished in two steps. In the first study, the five principles were operationalized by generating a questionnaire consisting of 45 items intended to capture the dimension of each principle. An exploratory maximum‐likelihood factor analysis indicated a relatively robust five factor structure that corresponded with the theoretical model. In the second study, the five factor model was validated using a confirmatory factor analysis. The model was then investigated using a two‐level linear mixed model with a random intercept. The results revealed that motivation and mathematical achievement were significant predictors of mathematical creativity.     

Some types of stability in the study of students' work values

It is suggested the concept of “stability” in the study of work values has at least three facets: (1) hierarchical stability, (2) response pattern stability, and (3) factor stability. These three facets are investigated using a sample of 165 Australian students, assessed using a work values list initially in Grade 12 and then $\text{2}\text{1}\text{2}$ years later during which time all had engaged in full-time tertiary study. Of the 13 work values assessed only independence and intellectual stimulation increased in importance and surroundings decreased. Canonical correlation analysis revealed four significant response patterns which indicated that the same work value may sometimes be stable or changeable in combination with other work values. It was also found using factor analysis that two rotated principal components appeared stable over time. They were identified as an Intellectual Concern factor and People Concern factor. An Extrinsic Concern factor for subjects as high school students was found to differentiate into a Status/Power Concern factor and a Co-workers Concern factor by the time of the second assessment. A Freedom Concern factor found in the Grade 12 assessment was not discovered until $\text{2}\text{1}\text{2}$ years later. The results confirm the need for further theoretical and methodological refinement of the notion of stability in the study of work values.     

Cognitive-affective stress propensity: A confirmatory factor analysis of stress propensity measures

A cognitive-affective stress propensity (CASP) construct is used to integrate previous findings regarding individual differences in experienced stress and resultant strain. A confirmatory factor analysis determined the structure of the CASP construct using a one factor model. This model and alternative models of factor structure contain two additional new constructs, psychological magnification and cognitive-affective connectivity. These two variables, and five trait variables previously identified as moderators in stress propensity research, were used in the analysis. Results show the one factor model to fit the data better than alternative models, thus supporting the notion of a latent construct (CASP).