The Family Adaptability and Cohesion Scale III in a Norwegian Sample

Olson's Family Adaptability and Cohesion Scales (FACES III) and Beavers' Self-report Family Inventory (SFI) were administered to 177 young Norwegian adults. Three questions were raised: (1) do the two-factor structure and curvilinear hypothesis of Olson's model of family functioning find support cross-culturally; (2) does curvilinearity depend on whether families of origin or procreation are described; and (3) does the data analytic technique dictate the results. In 1985, in their manual, Olson and his associates proposed the use of Distance from Center (DFC) scores to test the curvilinear hypothesis, a procedure that presupposes orthogonality between the dimensions of Cohesion and Adaptability, which was not found in several other samples nor in the present study. Using DFC scores, the curvilinear hypothesis received moderate support only when families of origin were described. A polynomic regression analysis gave clear evidence, however, of a linear relationship between the FACES dimensions and SFI Health Scores for both families of origin and families of procreation. The findings suggest that respondents do not have an implicit bipolar model of cohesion and adaptability.     

Predicted differences and differences between predictions

WhenK tests are given toN individuals, and for each individual there are two criterion measures, then (1) the multiple regression weight to be applied to the standard score for each test to predict the criterion-difference score equals the difference of the weights for predicting each criterion separately; (2) the difference between the predicted scores equals the predicted difference (each test being assigned the appropriate multiple regression weight); (3) the square of the multiple correlation between predicted and actual criterion-difference scores equals the sum of squares of the multiple correlations of the battery with each criterion less the product of these correlations and the correlation between predicted scores all divided by twice the quantity one minus the criterion intercorrelation; and (4) the variance of errors of estimating the criterion-difference score equals the sum of the variances of errors of estimating each criterion score minus twice the criterion intercorrelation, plus twice the correlation between predicted scores multiplied by the product of the square root of one minus the variance of errors of estimating one criterion and the corresponding square root for the second criterion.The author wishes to express his appreciation for the suggestions and guidance given by Dr. Harold Gulliksen in the preparation of this article. He also wishes to acknowledge the helpful comments of Dr. Paul Horst and Dr. Ledyard Tucker on certain phases of the development.     

Competitive state anxiety and self-confidence: intensity and direction as relative predictors of performance on a golf putting task

This study considered relationships between the intensity and directional aspects of competitive state anxiety as measured by the modified Competitive Sport Anxiety Inventory-2(D) (Jones & Swain, 1992) in a sample of 12 experienced male golfers. Anxiety and performance scores from identical putting tasks performed under three different anxiety-manipulated competitive conditions were used to assess both the predictions of Multidimensional Anxiety Theory (MAT; Martens et al., 1990) and the relative value of intensity and direction in explaining performance variance. A within-subjects regression analysis of the intra-individual data showed partial support for the three MAT hypotheses. Cognitive anxiety intensity demonstrated a negative linear relationship with performance, somatic anxiety intensity showed a curvilinear relationship with performance, and self-confidence intensity revealed a positive linear relation. Cognitive directional anxiety illustrated a positive linear relationship with putting performance. Multiple regression analyses indicated that direction (42% of variance) was a better predictor of performance than intensity (22%).     

Effects of Age,Domain, and Processing Demands on Memory Span: Evidence for Differential Decline

Analysis of cross-sectional data from the normative sample of the Wechsler Memory Scale - Third Edition (WMS-III) revealed different patterns of age-related differences in memory span measures depending on the type of memory item, processing demands, and the age of the older adult group. Regression of memory span on age revealed that the slope for Spatial Span raw scores was significantly more negative than the slope for Digit Span raw scores. There was no significant difference, however, either between the slopes for forward and backward Digit Span or between the slopes for forward and backward Spatial Span. Regression of Letter-Number Sequencing raw scores on age showed a distinctive, curvilinear pattern. Taken together, the present findings suggest that at least two mechanisms are involved in age-related differences in memory span. One mechanism, associated with a relatively linear decrease in memory span as a function of age, may differentially affect the storage of different types of information (e.g., sequences of digits vs. spatial locations). The other mechanism, evidenced by the curvilinear trend in Letter-Number Sequencing scores, may be tentatively attributed to a decline in executive aspects of working memory that becomes increasingly pronounced with age.     

A Computationally Efficient State Space Approach to Estimating Multilevel Regression Models and Multilevel Confirmatory Factor Models

Although the state space approach for estimating multilevel regression models has been well established for decades in the time series literature, it does not receive much attention from educational and psychological researchers. In this article, we (a) introduce the state space approach for estimating multilevel regression models and (b) extend the state space approach for estimating multilevel factor models. A brief outline of the state space formulation is provided and then state space forms for univariate and multivariate multilevel regression models, and a multilevel confirmatory factor model, are illustrated. The utility of the state space approach is demonstrated with either a simulated or real example for each multilevel model. It is concluded that the results from the state space approach are essentially identical to those from specialized multilevel regression modeling and structural equation modeling software. More importantly, the state space approach offers researchers a computationally more efficient alternative to fit multilevel regression models with a large number of Level 1 units within each Level 2 unit or a large number of observations on each subject in a longitudinal study.     

Estimating linear effects in ANOVA designs: The easy way

Research in cognitive science has documented numerous phenomena that are approximated by linear relationships. In the domain of numerical cognition, the use of linear regression for estimating linear effects (e.g., distance and SNARC effects) became common following Fias, Brysbaert, Geypens, and d'Ydewalle's (1996) study on the SNARC effect. While their work has become the model for analyzing linear effects in the field, it requires statistical analysis of individual participants and does not provide measures of the proportions of variability accounted for (cf. Lorch & Myers, 1990). In the present methodological note, using both the distance and SNARC effects as examples, we demonstrate how linear effects can be estimated in a simple way within the framework of repeated measures analysis of variance. This method allows for estimating effect sizes in terms of both slope and proportions of variability accounted for. Finally, we show that our method can easily be extended to estimate linear interaction effects, not just linear effects calculated as main effects.     

Evaluation of a networked self-testing program

The use of a computerized, multiple-choice test bank to present practice and assessment tests on a network was evaluated with 46 men and 119 women from a first-year class in psychology. A correlation of .65 (p < .001) between scores on a traditional paper-and-pencil test and scores on a computerized test provided some validity for the computerized assessment. Regression analysis showed that ability (previous academic performance) and motivation (number of practice tests taken) accounted for 73% of the explained variance in computerized test scores. Sex differences did not enter the regression equation significantly.     

概化理论缺失数据方差分量估计

各种心理调查、心理实验中, 数据的缺失随处可见。由于数据缺失, 给概化理论分析非平衡数据的方差分量带来一系列问题。基于概化理论框架下, 运用Matlab 7.0软件, 自编程序模拟产生随机双面交叉设计p×i×r缺失数据, 比较和探讨公式法、REML法、拆分法和MCMC法在估计各个方差分量上的性能优劣。结果表明：(1) MCMC方法估计随机双面交叉设计p×i×r缺失数据方差分量, 较其它3种方法表现出更强的优势; (2) 题目和评分者是缺失数据方差分量估计重要的影响因素。     

Adolescent Delinquency: Is Too Much or Too Little Parental Control a Problem?

Delinquent behavior is a parental concern during the period of adolescence. Previous theories and studies suggested that high parental control relates to lower delinquent behavior. However, Nye’s social control theory suggests a curvilinear rather than a linear relationship between parental control and delinquency. This study uses Nye’s social control theory to explore a curvilinear relationship between parental control and delinquency. Data concerning parental control and delinquency from Wave I of the National Longitudinal Study of Adolescent Health (Add Health) was analyzed using negative binomial regression in conjunction with “Svy” estimation method. The adolescents in the sample ranged from ages 12–21 years old (M = 15.55) and varied in race/ethnicity, family structure, and socioeconomic status. The quadratic term of parental control had a significant relationship with delinquency while controlling for the linear term of the parental control variable and covariates. The results suggested that high and low parental control related to higher levels of delinquent behavior. Moderate amounts of parental control related to lower levels of delinquent behavior. The practical and clinical implications of the findings are discussed within the study.     

Methodological pitfalls in predicting counseling success

Prediction models employing multiple linear regression of raw scores, multiple linear regression of factor scores, the single best predictor, and a nine-point decision rule index were compared. The subjects were 296 clients undergoing vocational counseling and evaluation. Predictor variables included performance ratings, demographic variables, and WAIS subtest scores; the criterion was employment status upon program completion. The least statistically sophisticated model, employing the single best predictor, was the most successful approach. Considerable shrinkage in power of prediction was demonstrated upon cross-validation particularly for multiple linear regression of raw scores model, indicating the necessity of cross-validating prediction schemes. Additional suggestions are made to those designing prediction studies.     

Order‐constrained linear optimization

Despite the fact that data and theories in the social, behavioural, and health sciences are often represented on an ordinal scale, there has been relatively little emphasis on modelling ordinal properties. The most common analytic framework used in psychological science is the general linear model, whose variants include ANOVA, MANOVA, and ordinary linear regression. While these methods are designed to provide the best fit to the metric properties of the data, they are not designed to maximally model ordinal properties. In this paper, we develop an order‐constrained linear least‐squares (OCLO) optimization algorithm that maximizes the linear least‐squares fit to the data conditional on maximizing the ordinal fit based on Kendall's τ. The algorithm builds on the maximum rank correlation estimator (Han, 1987, Journal of Econometrics, 35, 303) and the general monotone model (Dougherty & Thomas, 2012, Psychological Review, 119, 321). Analyses of simulated data indicate that when modelling data that adhere to the assumptions of ordinary least squares, OCLO shows minimal bias, little increase in variance, and almost no loss in out‐of‐sample predictive accuracy. In contrast, under conditions in which data include a small number of extreme scores (fat‐tailed distributions), OCLO shows less bias and variance, and substantially better out‐of‐sample predictive accuracy, even when the outliers are removed. We show that the advantages of OCLO over ordinary least squares in predicting new observations hold across a variety of scenarios in which researchers must decide to retain or eliminate extreme scores when fitting data.     

Support for a continuous (single-process) model of recognition memory and source memory

Does memory retrieval occur in a continuous or an all-or-none manner? The shape of the receiver operating characteristic (ROC) has been used to answer this question, with curvilinear and linear memory ROCs indicating continuous and all-or-none retrieval processes, respectively. Signal detection models (e.g., the unequal variance model) correspond to a continuous retrieval process, whereas threshold models (including the multinomial model and the recollection component of the dual-process model) correspond to an all-or-none process. In studies of source memory, Slotnick et al. (2000) and others have observed curvilinear ROCs (supporting the unequal variance model), whereas Yonelinas (1999) observed linear ROCs (supporting the dual-process model). We resolve these seemingly inconsistent results, showing that source memory ROCs are naturally curvilinear but can appear linear when nondiagnostic source information is included in the analysis. Furthermore, the unequal variance model accounted for both recognition memory and source memory ROCs, supporting a continuous process of memory retrieval.     

Between-teacher variance of students' teacher-rated risk for emotional,behavioral, and adaptive functioning

As schools increasingly adopt universal social, emotional, and behavioral screening, more research is needed to examine the effects of between-teacher differences due to error and bias on students' teacher-rated screening scores. The current study examined predictors of between-teacher differences in students' teacher-rated risk across one global and three narrow domains of behavioral functioning. Participants included 2450 students (52.1% male, 54.2% White) and 160 teachers (92.1% female, 80.3% White) from four elementary schools in one Southeastern U.S. school district. Teachers rated student behavior on the Behavior Assessment System for Children (Third Edition) Behavioral and Emotional Screening System (BESS)-Teacher Form and completed a survey about their training and perspectives of common behavior problems. Results of multilevel linear regression found between-teacher effects to be greater for internalizing risk scores (intraclass correlation = 0.23) than for externalizing risk scores (intraclass correlation = 0.12) or adaptive behavior scores (intraclass correlation = 0.14). Statistically significant student predictors in most models included student grade, gender, race and/or ethnicity, office discipline referrals, and course grades. We also detected effects of several teacher-level variables in one or more of the models, including teacher gender, teacher ratings of problem severity and concern for hypothetical children displaying behavior problems, and the covariance of random teacher intercept and teacher random slopes for students' office discipline referrals. Although these factors explained some teacher-level variance in students' risk scores, a notable amount of variance between teachers remains unexplained. Future research is needed to fully understand, reduce, and account for differences between teacher ratings due to error and bias.     

Latent variables should remain as such: Evidence from a Monte Carlo study

Use of subject scores as manifest variables to assess the relationship between latent variables produces attenuated estimates. This has been demonstrated for raw scores from classical test theory (CTT) and factor scores derived from factor analysis. Conclusions on scores have not been sufficiently extended to item response theory (IRT) theta estimates, which are still recommended for estimation of relationships between latent variables. This is because IRT estimates appear to have preferable properties compared to CTT, while structural equation modeling (SEM) is often advised as an alternative to scores for estimation of the relationship between latent variables. The present research evaluates the consequences of using subject scores as manifest variables in regression models to test the relationship between latent variables. Raw scores and three methods for obtaining theta estimates were used and compared to latent variable SEM modeling. A Monte Carlo study was designed by manipulating sample size, number of items, type of test, and magnitude of the correlation between latent variables. Results show that, despite the advantage of IRT models in other areas, estimates of the relationship between latent variables are always more accurate when SEM models are used. Recommendations are offered for applied researchers.     

Validity of the sex-role egalitarianism scale: Discriminating egalitarianism from feminism

To explore and to clarify the discriminant validity of the sex role egalitarianism construct, the Sex-Role Egalitarianism Scale (SRES) was administered in conjunction with the Attitudes Toward Women Scale (AWS) and eight personality trait measures to 108 individuals. Results indicated a hypothesized curvilinear relationship between scores on the SRES and the AWS. When personality traits were individually regressed on SRES and AWS, SRES scores uniquely accounted for variance in achievement locus of control, and AWS scores uniquely accounted for variance in need for autonomy and need for succorance. Findings were supportive of the viability of sex role egalitarianism as an attitudinal dimension having meaning beyond that of a traditional — feminist dimension.     

The relation of the reliability of multiple-choice tests to the distribution of item difficulties

Under certain assumptions an expression, in terms of item difficulties and intercorrelations, is derived for the curvilinear correlation of test score on the ability underlying the test, this ability being defined as the common factor of the item tetrachoric intercorrelations corrected for guessing. It is shown that this curvilinear correlation is equal to the square root of the test reliability. Numerical values for these curvilinear correlations are presented for a number of hypothetical tests, defined in terms of their item parameters. These numerical results indicate that the reliability and the curvilinear correlation will be maximized by (1) minimizing the variability of item difficulty and (2) making the level of item difficulty somewhat easier than the halfway point between a chance percentage of correct answers and 100 per cent correct answers.     

COST-BENEFIT CONSIDERATIONS IN CHOOSING AMONG CROSS-VALIDATION METHODS

There are two general methods of cross-validation: (a) empirical estimation, and (b) formula estimation. In choosing a specific cross-validation procedure, one should consider both costs (eg. inefficient use of available data in estimating regression parameters) and benefits (eg. accuracy in estimating population cross-validity). Empirical cross-validation methods involve significant costs, since they are typically laborious and wasteful of data, but under conditions represented in Monte Carlo studies, they are generally not more accurate than formula estimates. Consideration of costs and benefits suggests that empirical estimation methods are typically not worth the cost, except in a limited number of cases in which Monte Carlo sampling assumptions are not met in the derivation sample. Designs which use multiple samples to estimate the cross-validity of a single regression equation are clearly preferable to single-sample designs; the latter are never expected to be more accurate than formula estimates and thus are never worth the cost. Multi-equation designs are more accurate than single equation designs, but they appear to estimate the wrong parameter, and thus are difficult to interpret.     

Use of the family adaptability and cohesion evaluation scales in child clinical research

Although the FACES has become a widely used instrument for assessing children's family relations, the possible linearity vs. curvilinearity of its cohesion and adaptability scales has been treated inconsistently by investigators. This study evaluated whether samples of adolescent repeat offenders, young adult prisoners, and adolescent nonoffenders were discriminated better by a linear or curvilinear treatment of the FACES scores. Between-groups comparisons showed that significant effects were observed for each of the three curvilinear measures (cohesion-curvilinear, adaptability-curvilinear, distance-from-center) and only one of the two linear measures (cohesionlinear). These findings support the superiority of a curvilinear treatment of the FACES. In light of these findings, it is suggested that investigators who use FACES evaluate the linearity of the scales and determine whether a linear or curvilinear treatment of the data produces more meaningful results.     

A theoretical distribution for mental test scores

The negative hypergeometric distribution of raw scores on mental tests is derived from certain assumptions relating to test theory. This result is checked empirically in a number of examples. Further derivations lead to the bivariate distribution of parallel tests which is also verified with actual data. The bivariate distribution of raw score and true score is also derived from a further assumption. This distribution is used to set confidence limits for true scores for persons with a given raw score.This work was supported in part by contract Nonr-2752(00) between the Office of Naval Research and Educational Testing Service. Reproduction in whole or in part for any purpose of the United States Government is permitted.