Wei Ming Luh

Reliability and validity of the Chinese version of the Job Content Questionnaire in Taiwanese workers.

The reliability and validity of four selected scales—namely, job control, psychological demands, supervisor support, and coworker support—from the Chinese Version of the Job Content Questionnaire (C—JCL) were studied in 551 male and 648 female workers in Taiwan. Cronbach’s alpha coefficients for job control, supervisor support, and coworker support were all above .80. Whereas it was .55 for psychological demands, indicating insufficient internal consistency for this subscale. Participants responded to questionnaire items relatively consistently over a 3-month period. Exploratory factor analyses disclosed 4 empirical factors, which corresponded closely with theoretical constructsofthe JCL.As predicted by the model, lowest levels of job satisfaction were found in workers in “iso-strain” category (i.e., high demands combined with low control and low social support). These findings indicated that the C—JCL is reliable and valid for assessing psychosocial work conditions among Taiwanese workers, although further improvement is needed for the psychological demands subscale.

Context availability and the recall of abstract and concrete words

Predictions of an automatic-imagery, strategic-imagery, and context-availability hypothesis of concreteness effects in free recall were examined. In each experiment, recall of abstract and concrete words controlled for rated context availability was compared with the typical situation in which context availability is confounded with imageability. In Experiment 1, a directed intentional-recall task produced concreteness effects in recall. Experiment 2 compared concreteness effects in recall following three orienting tasks: imagery rating, context-availability rating, and a directed intentional-memory task. Concreteness effects in the context-availability-controlled condition were found following the imagery-rating and the directed intentional-memory tasks, but not after the context-availability-rating task. In Experiment 3, subjects reported the strategies that they used to encode the list. Subjects reporting an imagery strategy showed concreteness effects for words controlled for rated context availability, but those not reporting it did not. These results support a strategic-imagery view of concreteness effects in free recall.

A powerful transformation trimmed mean method for one-way fixed effects ANOVA model under non-normality and inequality of variances

The present study proposes a transformation trimmed mean method for the one-way e xed effects ANOVA model to deal with non-normality and inequality of variances. First, the data are trimmed on both sides. Then the test statistic is transformed by Johnson’ s method for each group to deal with non-normality. Finally, an approximate test such as the Alexander‐ Govern test, the James second-order test or the Welch test is applied to test the equality of population trimmed means. Monte Carlo simulation results show that, under non-normal cases, the proposed method can control Type I error rates, and has greater power than other competitors, such as the original Alexander‐ Govern method, the Alexander‐ Govern trimmed mean method, the James second-order trimmed mean method, the Welch trimmed mean method, as well as the corresponding bootstrapping methods. The Johnson’ s transformation trimmed mean method can be programmed easily and is highly recommended as an alternative for the one-way e xed effects ANOVA model if normality or equality of variances cannot be assumed.

An invertible transformation two-sample trimmed t-statistic under heterogeneity and nonnormality

As various transformation methods are unsatisfactory to control Type-I error rate under heterogeneity and nonnormality for the two-sample case, an invertible transformation trimmed t-statistic is proposed, and the confidence interval is derived. Simulation results and experimental data are also presented.

Factors Related to Depression during Menopause: A Study in Southern Taiwan

The purpose of this study was to survey the prevalence of womens depression during menopause and investigate the relationships between menopause depression and variables including self-concept, menopause attitudes, life-event stresses, climacteric physiological symptoms, and demographics. Research was conducted using a cross-sectional study and self-administered questionnaires. We randomly selected female participants between the ages of 45 and 55 years from Tainan City in southern Taiwan using the cluster method. Two hundred sixty-six valid questionnaires were analyzed. Results were as follows: (a) The prevalence of depression among participants was 31.2%; (b) depression was positively associated with life-event stresses and climacteric physiological symptoms (r =.23 and .34, respectively) and was negatively associated with attitude toward menopause and self-concept (r = −.31 and −.65, respectively); and (c) by logistic regression analyses, climacteric physiological symptoms and self-concept were associated with menopause depression. This preliminary survey suggests that low self-concept and increased climacteric physiological symptoms are closely related to depression among menopausal women. Larger scale follow-up studies are warranted to investigate factors that predispose women to depression during menopause.

Using Johnson's transformation and robust estimators with heteroscedastic test statistics: An examination of the effects of non-normality and heterogeneity in the non-orthogonal two-way ANOVA design

The present study proposes a procedure that combines Johnsons transformation and the trimmed means method to deal with the problem of non-normality. An approximate test such as the Alexander-Govern test or Welch-James type test is then employed to deal with the heterogeneity of cell variance in the non-orthogonal two-way fixed effects completely randomized design. Both unweighted and weighted means analyses are considered. The empirical Type I error rates and the statistical power for comparing population means are investigated by Monte Carlo simulation. The simulated results show that Johnsons transformation with trimmed mean and the approximate test is valid in terms of Type I error rate control, and that the magnitude of the statistical power for non-normal distributions is better than that of conventional methods.

Approximate sample size formulas for the two-sample trimmed mean test with unequal variances

Yuens two-sample trimmed mean test statistic is one of the most robust methods to apply when variances are heterogeneous. The present study develops formulas for the sample size required for the test. The formulas are applicable for the cases of unequal variances, non-normality and unequal sample sizes. Given the specified alpha and the power (1-beta), the minimum sample size needed by the proposed formulas under various conditions is less than is given by the conventional formulas. Moreover, given a specified size of sample calculated by the proposed formulas, simulation results show that Yuens test can achieve statistical power which is generally superior to that of the approximate t test. A numerical example is provided.

Approximate transformation trimmed mean methods to the test of simple linear regression slope equality

To deal with the problem of non-normality and heteroscedasticity, the current study proposes applying approximate transformation trimmed mean methods to the test of simple linear regression slope equality. The distribution-free slope estimates are first trimmed on both sides and then the test statistic t is transformed by Johnsons method for each group to correct non-normality. Lastly, an approximate test such as the James second-order test, the Welch test, or the DeShon-Alexander test, which are robust for heterogeneous variances, is applied to test the equality of regression slopes. Bootstrap methods and Monte Carlo simulation results show that the proposed methods provide protection against both unusual y values, as well as unusual x values. The new methods are valid alternatives for testing the simple linear regression slopes when heteroscedastic variances and nonnormality are present.

Optimum sample size allocation to minimize cost or maximize power for the two-sample trimmed mean test.

When planning a study, sample size determination is one of the most important tasks facing the researcher. The size will depend on the purpose of the study, the cost limitations, and the nature of the data. By specifying the standard deviation ratio and/or the sample size ratio, the present study considers the problem of heterogeneous variances and non-normality for Yuens two-group test and develops sample size formulas to minimize the total cost or maximize the power of the test. For a given power, the sample size allocation ratio can be manipulated so that the proposed formulas can minimize the total cost, the total sample size, or the sum of total sample size and total cost. On the other hand, for a given total cost, the optimum sample size allocation ratio can maximize the statistical power of the test. After the sample size is determined, the present simulation applies Yuens test to the sample generated, and then the procedure is validated in terms of Type I errors and power. Simulation results show that the proposed formulas can control Type I errors and achieve the desired power under the various conditions specified. Finally, the implications for determining sample sizes in experimental studies and future research are discussed.

The structure of a web-based HIV testing belief inventory (wHITBI) for college students: The evidence of construct validation

The aim of this study is to develop a web-based HIV testing belief Inventory (wHITBI) and to examine the reliability and validity of the scores of the instrument. Steps for developing the item pool are described. Students from one major university in the south-eastern US were recruited. The sample (n = 440) was randomly split into equivalent halves for an exploratory factor analysis (EFA) and confirmatory factor analysis (CFA). The EFA showed that the wHITBI items were loaded in a way consistent with the four theoretical constructs: perceived benefits, concerns of HIV risk, perceived stigma, and availability/accessibility. CFA using LISREL 8.7 confirmed the structure of the inventory indicated by good model-fit indices (χ2/d.f. = 1.79; root mean square error of approximation = .06; non-normed fit index = .89; comparative fit index = .90; incremental fit index = .91; root mean square = .07), with all factors loaded significantly (p < .001). The composite reliability ranged from .67 to .76. The results also showed that these scales can distinguish well among HIV-tested and never-tested students. The validated wHITBI has implications on future applications for the development and evaluation of HIV prevention and education programs delivered via the Internet.