Search

When data are available on multiple individuals measured at multiple time points that may vary in number or inter-measurement interval, hierarchical linear models (HLM) may be an ideal option. The present paper offers an applied tutorial on the use of HLM for developing growth curves depicting natural changes over time. We illustrate these methods with an example of body mass index (BMI; kg/m(2)) among overweight and obese adults. We modelled among-person variation in BMI growth curves as a function of subjects' baseline characteristics. Specifically, growth curves were modelled with two-level observations, where the first level was each time point of measurement within each individual and the second level was each individual. Four longitudinal databases with measured weight and height met the inclusion criteria and were pooled for analysis: the Framingham Heart Study (FHS); the Multiple Risk Factor Intervention Trial (MRFIT); the National Health and Nutritional Examination Survey I (NHANES-I) and its follow-up study; and the Tecumseh Mortality Follow-up Study (TMFS). Results indicated that significant quadratic patterns of the BMI growth trajectory depend primarily upon a combination of age and baseline BMI. Specifically, BMI tends to increase with time for younger people with relatively moderate obesity (25 BMI

Background: This study was based on over 30,000 respondents who completed General Social Surveys between 1978 and 2002. Aims: We approached these respondents prospectively, comparing and contrasting the responses of those who subsequently died by suicide (N = 141) with those who died from all other causes (N = 9,115). Method: We employed chi-square and logistic regression analyses of important demographic confounders to test for statistically significant differences between suicide decedents and all other decedents. Results: Suicide decedents died on average 2 years sooner than all other decedents. When covariates of age and gender were applied, suicide decedents exhibited greater acceptance of suicide for dealing with various adverse life circumstances, were more likely to have been the gun owners in their households, lived in regions where gun ownership was more commonplace, and held less strong religious beliefs and less of a belief of an afterlife. Conclusion: The observed affinity between attitudes of suicide acceptability and completed suicide suggests a potential for creating a meaningful assessment tool to identify those positioned at the extreme end of the suicide risk continuum.

The procedures for interpreting individual test scores are often given incorrectly in many of the applied texts pertaining to tests and measurements. The issues in choosing the correct standard error of estimate for a particular score interpretation problem are discussed in detail using the basic concepts of Classical Test Theory. Examples are provided to illustrate how a test user can take into account the reliability of a test when making recommendations for students based on a single test score and interpreting different scores resulting from retesting with a standardized test. The issues in computing these measures by hand are also discussed. The intended audience for this paper is the test user and student who have only a basic knowledge of psychometric theory and statistics.

The development and psychometric properties of the Extended Satisfaction With Life Scale (ESWLS) are described in detail. The ESWLS is a 50-item self-report scale that measures satisfaction with life in nine domains. It can be completed by most people in under 20 minutes and can be used by researchers and clinicians. The readability of the ESWLS was estimated to be between the seventh and tenth grade levels. Internal consistency, estimated by coefficient α, ranged from 0.81 to 0.96 for the individual subscales. Two-week test-retest reliability coefficients ranged from 0.74 to 0.87. Results of exploratory and confirmatory factor analyses strongly supported the factor structure of the ESWLS. Preliminary evidence of convergent and discriminant validity is provided as well as preliminary norms.

POWPAL is a simple, interactive computer program that computes effect sizes and power estimates from summary statistics. Additionally, the program provides estimates of future sample sizes needed for various combinations of power, significance levels, and effect sizes.

Unicorn, a BASIC computer program that automates the Box-Cox transformation, is described. Unicorn reads free field ASCII data files, is completely interactive, and is user friendly. Unicorn can handle files with or without missing data. Descriptive statistics and tests of normality are provided for each variable in untransformed format. Statistics include the first four moments of the distribution, the third and fourth L-moments, the Kolmogorov-Smirnov test, and D'Agostino et al.'s tests of normality. The user is then offered a choice of minimizing skewness or departures from normality overall. The optimal transformation is then found and descriptive statistics and tests of normality are provided for the transformed data. A new ASCII data file is written with transformed data. This new data file is then available for further analysis.

Forty-two experimental studies based on the Dunn and Dunn Learning Style Model and conducted between 1980–1990 were identified to determine the value of teaching students through their learning-style preferences. The studies were rated according to Lytton and Romney's (1991) Quality Rating Scales. A jury determined that, of the 42 studies, 6 studies evidenced serious threats to validity. The 36 remaining studies provided a database of 3,181 participants. Results were synthesized through meta-analysis. Eight variables coded for each study produced 65 individual effect sizes. The overall, unweighted group effect size value (r) was .384, and the weighted effect size value was .353 with a mean difference (d) of 755. Referring to the standard normal curve, this suggests that students whose learning styles are accommodated would be expected to achieve 75% of a standard deviation higher than students who have not had their learning styles accommodated. This finding indicates that matching students' lear...

Analysis of change points in psychotherapy process could increase our understanding of mechanisms of change. In particular, naturalistic change point detection methods that identify turning points or breakpoints in time series data could enhance our ability to identify and study alliance ruptures and resolutions. This paper presents four categories of statistical methods for detecting change points in psychotherapy process: criterion-based methods, control chart methods, partitioning methods, and regression methods. Each method's utility for identifying shifts in the alliance is illustrated using a case example from the Beth Israel Psychotherapy Research program. Advantages and disadvantages of the various methods are discussed.

Comorbidity with disruptive behavior disorders may have important implications for exposure-based cognitive behavioral treatments of children with OCD. Child noncompliance and parent-child conflict may interfere with performance of exposure activities and completion of therapeutic homework assignments, thus diminishing response to treatment. We investigated whether response to exposure and response prevention (ERP) can be enhanced if disruptive behavior is treated first with parent management training (PMT). A multiple-baseline across-responses design was used to investigate the effects of ERP with or without PMT in six children (age range 9-14 years) with OCD and disruptive behavior. Weekly ratings of OCD were conducted for four weeks to establish baseline. After that, children were randomly assigned to receive six weekly sessions of PMT and then twelve weekly sessions of ERP (ERP-plus-PMT condition) or to receive ERP after a six week waiting period (ERP-only condition). The outcome assessments were conducted weekly using the Child Yale-Brown Obsessive Compulsive Scale (CY-BOCS) administered by an experienced clinician, who was blind to treatment assignment. Three subjects in the ERP-plus-PMT condition evidenced a 39 percent reduction in the CY-BOCS score versus a 10 percent reduction in three subjects in the ERP-only condition. The results of our single-subject study suggest the feasibility and positive effects of combining ERP with PMT for children with OCD complicated by disruptive behavior.

Scruggs and Mastropieri (Behaviour Research and Therapy, 32, 879–883, 1994) take issue with criticisms of their PND (Percent of Nonoverlapping Data) statistic that we offered in our recent article (Allison & German, Behaviour Research and Therapy, 31, 621–631, 1993), which advocated a regressionbased method for obtaining effect sizes in single-subject studies. They contend that their PND approach has several advantages over our approach because: (1) they believe that, unlike ours, it can take advantage of the small number of observations that are typically available in single-case studies; (2) it is simple to compute; (3) it frees researchers from traditional regression assumptions of normality, homogeneity of variance, and independence of observations and residuals; and (4) it correlates with visual judgements made by experts. As we shall argue, these claims are built upon very questionable assumptions and they are very difficult to substantiate. In addition, we show that the expected value of the PND is so strongly related to sample size as to be rendered meaningless.

The standard deviation is the most commonly reported measure of variability for a group of measurements. It is used as a descriptive statistic to indicate the degree of homogeneity or heterogeneity present in a given group of measurements and as an estimate of a population parameter. It is also used as a component of formulas for other statistics such as the standard error of estimate and the standard error of measurement. This article discusses issues in teaching students about computing the standard deviation in both of these contexts. The standard deviation does not identify the same quantity in all of these situations.

POWCOR is a BASIC program that computes power or needed sample sizes for detecting the differences between pairs of correlated or independent correlation coefficients. The program can compute power for unequal samples sizes, given independent correlations.

Previous research [Campbell and Muncer: Journal for the Theory of Social Behavior 17: 489–512, 1987; Campbell et al.: Aggressive Behavior 18: 95–108, 1992] suggests that men and women hold different social representations or implicit theories of their own aggression. Men view it is an instrumental act (a means of obtaining and exercising power to gain social rewards), while women view it an expressive act (a cathartic discharge of anger). In the present study, communal/agentic personality styles and gender identity are examined as possible mediators of the relation between sex and Expaag–a psychometric measure of adherence to an expressive representation of aggression. In addition a measure of self-reported aggression is included. The highest correlation appeared between sex and Expaag. Gender identity and interpersonal style made no significant improvement in explained variance in a multiple regression analysis after sex had been entered. An instrumental social representation of aggression was significantly and positively correlated with number of reported aggressive acts. © 1993 Wiley-Liss, Inc.

A subscale of the Three-Factor Eating Questionnaire (TFEQ), a subscale of the Dutch Eating Behavior Questionnaire (DEBQ), and the Revised Restraint Scale (RS) were administered to 90 1 undergraduates. Test-retest reliability (on 34 subjects) was highest for the RS (r =.95) and roughly equal for the DEBQ (r =.92) and the TFEQ (r =.91). Internal consistency was highest for the DEBQ (α =.95), moderate for the TFEQ (α =.90), and lowest for the RS (α =.82). The DEBQ was the most homogeneous scale, with a single principal component accounting for 68.2% of the variance

With survey data collected primarily from peer support group participants, the authors compared stigmatization responses of 462 parents losing children to suicide with 54 other traumatic death survivors and 24 child natural death survivors. Parents who encountered harmful responses and strained relations with family members and non-kin reported heightened grief difficulties. After controlling for time since the death and whether a child's death was traumatic or not, stigmatization continued to be associated with grief difficulties, depression, and suicidal thinking. Suicide survivors reported little differences in stigmatization from other-traumatic-death survivors, a result consistent with other recent studies, suggesting more convergence between these two populations than divergence.

Etude de 24 incidents agressifs par les «offenseurs». Ces offenseurs doivent trier les incidents en fonction de ce qui est plus naturel, plus logique ou plus agreable pour eux. Les resultats sont classes en termes de schema cognitif sous-jacent

We examined the relevance of the level and pattern of the therapeutic alliance in 44 cases of three different, manualized 30-session treatments using patient ratings of the Working Alliance Inventory after each session. It was hypothesized that both high-alliance level and either a linear increase in alliance rating or a series of brief rupture-and-repair episodes would be found in successful treatments. We also hypothesized that a more global high-low-high pattern predicted in the literature would not be present. Consistent with the literature, higher alliance levels were found to be related to improved outcome. As predicted, we did not find a global, high-low-high pattern. Local rupture-and-repair patterns were found in 50% of the cases; linear trends were found in 66% of the cases. There was no relationship between outcome and either pattern. We found no differences among the treatments.

Although many clinical studies document a relationship between gambling and suicidality, evidence of this association in general population surveys has been mixed. Probing this association in a nationally representative sample of young adults with data from the National Longitudinal Survey of Adolescent Health, we made same gender comparisons of depression and suicidality between 298 at-risk gamblers and 13,000 others. Although gamblers of both genders showed higher depression, only females reported significantly higher suicide thoughts and attempts. Males with gambling problems were no more likely than nongamblers to have suicide thoughts or to make prior suicide attempts on three separate measurement occasions.

Motivation In cluster analysis, the validity of specific solutions, algorithms, and procedures present significant challenges because there is no null hypothesis to test and no 'right answer'. It has been noted that a replicable classification is not necessarily a useful one, but a useful one that characterizes some aspect of the population must be replicable. By replicable we mean reproducible across multiple samplings from the same population. Methodologists have suggested that the validity of clustering methods should be based on classifications that yield reproducible findings beyond chance levels. We used this approach to determine the performance of commonly used clustering algorithms and the degree of replicability achieved using several microarray datasets. Methods We considered four commonly used iterative partitioning algorithms (Self Organizing Maps (SOM), K-means, Clutsering LARge Applications (CLARA), and Fuzzy C-means) and evaluated their performances on 37 microarray datasets, with sample sizes ranging from 12 to 172. We assessed reproducibility of the clustering algorithm by measuring the strength of relationship between clustering outputs of subsamples of 37 datasets. Cluster stability was quantified using Cramer's v 2 from a kXk table. Cramer's v 2 is equivalent to the squared canonical correlation coefficient between two sets of nominal variables. Potential scores range from 0 to 1, with 1 denoting perfect reproducibility. Results All four clustering routines show increased stability with larger sample sizes. K-means and SOM showed a gradual increase in stability with increasing sample size. CLARA and Fuzzy C-means, however, yielded low stability scores until sample sizes approached 30 and then gradually increased thereafter. Average stability never exceeded 0.55 for the four clustering routines, even at a sample size of 50. These findings suggest several plausible scenarios: (1) microarray datasets lack natural clustering structure thereby producing low stability scores on all four methods; (2) the algorithms studied do not produce reliable results and/or (3) sample sizes typically used in microarray research may be too small to support derivation of reliable clustering results. Further research should be directed towards evaluating stability performances of more clustering algorithms on more datasets specially having larger sample sizes with larger numbers of clusters considered.

To investigate the prevalence of thought disorder and the possible appropriateness of the Bizarre–Idiosyncratic Thinking Scale for children and adolescents with severe psychiatric disorders, 96 child and adolescent inpatients and day hospital patients, ages 6 to 18 years, at a state psychiatric hospital were rated by review of retrospective records using Marengo and Harrow's 1986 Evaluation of Bizarre–Idiosyncratic Thinking Scale for the presence of thought disorder in the Thematic Apperception Test and Rorschach Inkblot Test responses. Although the Evaluation of Bizarre–Idiosyncratic Thinking Scale had not been previously used with children and adolescents, the analysis suggested possible indications of thought disorder in several diagnostic groups. No significant differences were found on the Rorschach between patients with Schizophrenia and Psychosis, Not Otherwise Specified and those with Attention Deficit Hyperactivity Disorder, Major Depression, and Conduct Disorder. On the basis of the Thinking Scale ratings, the Thematic Apperception Test responses showed significantly higher ratings of thought disorder for patients with Schizophrenia and Psychosis, Not Otherwise Specified. There was no general relation between thought disorder and age or IQ, but schizophrenic patients, aged 13 and older, had more thought disorder than schizophrenic patients who were younger than 13.

MCA.EXE is a program that performs multiple correspondence analysis of categorical data on MS/DOS and PC/DOS computers. Using the formulas provided by Carroll and Green, the program also generates matrices for further INDSCAL analysis.

Using the 1987 General Social Survey data, the factors related to the disproportionately higher rates of cigarette smoking by Black Americans were investigated. Previous studies have found smoking to be highly correlated with age, social class, and occupational stress, among other factors. It is uncertain whether race is an independent predictor of smoking rates or whether it is primarily a correlate of other demographic variables. Loglinear modeling and logit regression analysis suggested that racial differences between Blacks and Whites in smoking may be spurious. The multivariate analysis established that class and stress differences remained more potent than race in accounting for variations in smoking behavior.

Summary Delay of gratification and temporal span were investigated in 58 girls, ages four to nine years, by estimated story durations, delay of gratification in a direct choice money test, and reproduction of a 60-second interval. All measures were highly related (p > .001) to age and cognitive-temporal stage, with the former relationships of greater magnitude.

Factor and cluster analyses are distinctly different multivariate procedures with different goals. However, when used in a complementary fashion, each set of methods can be used to enhance the interpretation of results found in the other set of methods. Simple examples illustrating the joint use of the methods are provided.

Rigidity vs flexibility has often been mentioned in discussions of creativity and problem solving. The present study investigated the relation of a non-verbal test of rigidity (Breskin Rigidity Test) to tests of semantic redefinition, associational fluency, inductive reasoning, and drawing completion. The performance of flexible Ss was significantly better than the performance of rigid Ss on all tests but the associational fluency test.

This study investigated the degree to which 43 college Ss could decode “sentograms,” a form of graphic expression of emotions discussed by Clynes. Ss could decode sentograms at better than chance levels. However, accuracy in matching sentograms to labels of specific emotions was considerably worse than in matching sentograms to broader categories of affects.

Spearman's rho is a special case of the product-moment r. As such, a principal components analysis of inter-rater rhos among k raters on n cases may be performed. The size of the first principal component is proportional to the amount of inter-rater agreement and can be equated to Kendall's W. The principal components procedure can also reveal the presence of differing rating policies and can isolate subgroups of raters.

The BMD12V MANOVA program does not compute the appropriate between-subjects tests in mixed effects designs. Although James (1979) recently suggested a technique for obtaining between-subjects tests, his formulas are shown to be incorrect. Suggestions are given for obtaining the appropriate statistics.

Summary.-Warehime and Jones found moderate correlations between social desirability and short-term mood level scores on the Wessman-Ricks Personal Feeling Scales. They suggested that it might be valuable to study parterns of dissimulation in self-reports of immediate feeling states. In the present authors' studies involving long-term mood reports and multivariate methods, social desirability was shown to have a moderately weak relationship with measures of mood lwel and mood variability. Under conditions of confidentiality and good rapport, repeated self-reports of mood appeared to share more variance with important personality characteristics and situational influences than with social desirability per se. warehime and Jones (1972) reported several moderately positive but significant correlations among the Edwards (1957) Social Desirability Scale, the Marlowe-Crowne Social Desirability Scale (Crowne & Marlowe, 1964), and mood level scores on the Wessman and Ricks (1966) Personal Feeling Scales. Relevant findings have been obtained in our own studies where the MMPI L and K scales, the Crowne-Marlowe scale, and various measures of defensiveness showed slight to moderate correlations with avowals of high positive mood levels on some Personal Feeling subscales (Wessman & Ricks, 1966; Gorman & Wessman, 1974). Warehime and Jones concluded their note with the statement that "Results provide suggestive leads for investigation of patterns of dissimulation in self-report of immediate feeling states." Through our research with the Personal Feeling Scale, we have been able to examine some of the relationships of social desirability and dissimulation to subscale scores and feel that we can provide a partial answer to the very complex issue raised by Warehime and Jones (1972). In this note, we attempt to show that the mood scales of the Personal Feeling Scale seem to be good measures of affect and when well administered, appear to be only slightly influenced by social desirability and dissimulation. In our studies, Ss either responded anonymously (Gorman & Wessrnan, 1974) or had established considerable rapport with the investigators over a 3-yr. period (Wessman & Ricks, 1966). In these situations, in which the press toward dissimulation was presumably low, social desirability scores showed some slight correlations with some mood level scores and no correlations with others. Those few significant correlations (none higher than .28) indicated that Ss scoring high on social desirability reported higher levels of harmony on the harmony-anger measures. Clearly, happiness, energy, self-confidence, and other moods measured by the Personal Feeling Scale are desirable qualities, but

Three groups of college drug users--those expressing a preference for marijuana, amphetamines, or barbiturates--reported on experiences resulting from their drugs, and these groups, together with a group of nondrug-users, completed the Omnibus Personality Inventory (OPI). On the OPI the nonusers appeared more conventional and conforming than each drug group. Those preferring marijuana were less anxious than any other group, but the most distinctive group was that preferring barbiturates, which appeared the least intellectually inclined and the most emotionally distressed. This group's personality dispositions also seemed generally concordant with the experiences they reported from the drug.

Some years back we began an inquiry that sometimes proved rewarding, yet more often has filled us with dissatisfaction, uncertainty, and a tormenting sense of the many questions still to be asked and answered. Our quest was one that has lured many others throughout the ages, namely, a search to understand the meaning of time and its relationships to other aspects of behavior and experience.

As locus of control involves generalized expectancies regarding the determination of events, it should relate to temporal attitudes and experiences. Correlational data from 89 subjects on the Rotter-Mirels Locus of Control and Ricks-Epley-Wessman Temporal Experience Questionnaire scales supported the hypothesis that the reported time experiences and orientations of external control subjects would be significantly more harassed and pressured, ciscontinuous and undirected, procrastinating and inefficient, and inconsistent and changeable than those of internal control subjects. Discussion forcuses on the self-defeating cycle of disorganization and victimization experienced by individuals with an external locus of control.

1 The Emergence of Human Awareness and Concepts of Time.- 2 Toward a Dialectical Interpretation of Time and Change.- 3 The Temporal Transition from Being Together to Being Alone: The Significance and Structure of Children's Bedtime Stories.- 4 Perception and Concept of Time: A Developmental Perspective.- 5 The Time of Youth.- 6 Memories of Tomorrow: On the Interpenetrations of Time in Later Life.- 7 Images, Values, and Concepts of Time in Psychological Research.- 8 The Modern Consciousness and the Winged Chariot.

Summary.-A group of 64 undergraduates (39 women, 25 men) was given Zuckerman's Sensation-Seeking Scale and Cattell's 16 PF. High sensation-seekers were characterized by grearer dominance, surgency, adventurousness, suspicion, and bohemian unconcernedness and by less shrewdness and self-sentiment control than low sensation-seekers. This pattern was more pronounced for women than men. The relationship of chis constellation to psychopathy was discussed. The Sensation-Seeking Scale (SSS) was developed by Zuckerman, Kolin, Price, and Zoob ( 1964) as a measure of individual differences in the need for stimulus variability. Since its development, the SSS has been shown to correlate positively with the MMPI. F, Pd, Ma, and Im scales and negatively with the MMPI K, L, and R scales (Zuckerman & Link, 1968; Blackburn, 1969). It has been recently suggested by Blackburn (1969) that the SSS might be a measure of psychopathic personality traits and a measure of psychotic symptomatology. The SSS, however, has also been shown to correlate positively with measures that are essentially independent of pathology, such as field independence (Zuckerman, et al., 1964; Zuckerman 81 Link, 1968), extraversion (Farley S: Farley, 1967), and the n autonomy, n change, and n exhibitionism scales of both the Edwards PPS and the Zuckerman Adjective Checklist (Zuckerman & Link, 1968). Negative correlations obtained for the SSS and measures of n affiliation, n orderliness, n nurmrance, and n deference (Zuckerman & Link, 1968). The present study investigated the source trait correlates of the SSS as measured by Cattell's Sixteen Personality Factor Test (16 PF; Cattell & Eber, 1967). The 16 PF was chosen because it has been repeatedly shown to sample n broad spectrum of traits in both the pathological and nonpachological domains (Hundleby & Connor, 1968; Catcell & Bolton, 1969). It was predicted that while the SSS would correlate with those 16 PF traits which discriminate between psychopaths and non-psychopaths, it would also correlate with nonpathological traits which discriminate between Ss who prefer active, chaotic, and ueative modes of behavior and Ss who prefer more subdued and ordered modes. Sixty-four undergraduates (39 women, 25 men) who were enrolled in introductory psychology courses at Nassau Community College were given the 22-item male-female form of the SSS (Zuckerman, et al., 1964) and Form A of the 1967 edition of the 16 PF (Cattell & Eber, 1967). All Ss anonymously answered both questionnaires. Table 1 presents the product-moment correlations of the SSS with 16 PF variables for women, for men, and for the combined group. Although the men's data did not yield the significant correlations noted for women, the pattern of correlations for the two sets of data were similar, suggesting that larger samples might provide more meaningful patterns of correlates. The factor pattern Ef, G-, Lf, Mf, QR- of the high SSS scorers is also

Ss were asked to indicate points 1 week, 7 months, 3 years, and 9 years in the past and future on two time lines representing birth to present and present to death. Data for 90 college-age Ss fit a psychophysical power function following Stevens’s law. with negatively accelerated growth indicating proportionately greater linear representation of periods nearer to the present. Variability was greater for the representations of the future than of the past, with monotonic increases in variability as distance from the present increased.

(1968). Social Desirability Factors and the Eysenck Personality Inventory. The Journal of Psychology: Vol. 69, No. 1, pp. 75-83.

This article introduces the Interpersonal Guilt Rating Scale-15 (IGRS-15), a brief clinician-rated tool for the clinical assessment of interpersonal guilt as conceived in Control-Mastery Theory (CMT; Silberschatz, 2015; Weiss, 1993), and its psychometric proprieties. The items of the IGRS-15 were derived from the CMT clinical and empirical literature about guilt, and from the authors' clinical experiences. Twenty-eight clinicians assessed 154 patients with the IGRS-15, the patient self-reported Interpersonal Guilt Questionnaire-67 (IGQ-67; O'Connor, Berry, Weiss, Bush, & Sampson, 1997), and the Clinical Data Form (CDF; Westen & Shedler, 1999). A semi-exploratory factor analysis pointed to a four-factor solution in line with the kinds of guilt described in CMT: Survivor guilt, Separation/disloyalty guilt, Omnipotent responsibility guilt, and Self-hate. The test-retest reliability of the IGRS-15 was good. Moreover, the IGRS-15 showed good concurrent and discriminant validity with the IGQ-67. IGRS-15 represents a first step in the direction of supporting the clinical judgment about interpersonal guilt with an empirically sound and easy-to-use tool.