Li-tze Hu’s research while affiliated with University of California, Santa Cruz and other places

A longitudinal study of 1st-year university student adjustment examined the effects of academic self-efficacy and optimism on students' academic performance, stress, health, and commitment to remain in school. Predictor variables (high school grade-point average, academic self-efficacy, and optimism) and moderator variables (academic expectations and self-perceived coping ability) were measured at the end of the first academic quarter and were related to classroom performance, personal adjustment, stress, and health, measured at the end of the school year. Academic self-efficacy and optimism were strongly related to performance and adjustment, both directly on academic performance and indirectly through expectations and coping perceptions (challenge-threat evaluations) on classroom performance, stress, health, and overall satisfaction and commitment to remain in school. Observed relationships corresponded closely to the hypothesized model. (PsycINFO Database Record (c) 2012 APA, all rights reserved)

This article examines the adequacy of the “rules of thumb” conventional cutoff criteria and several new alternatives for various fit indexes used to evaluate model fit in practice. Using a 2‐index presentation strategy, which includes using the maximum likelihood (ML)‐based standardized root mean squared residual (SRMR) and supplementing it with either Tucker‐Lewis Index (TLI), Bollen's (1989) Fit Index (BL89), Relative Noncentrality Index (RNI), Comparative Fit Index (CFI), Gamma Hat, McDonald's Centrality Index (Mc), or root mean squared error of approximation (RMSEA), various combinations of cutoff values from selected ranges of cutoff criteria for the ML‐based SRMR and a given supplemental fit index were used to calculate rejection rates for various types of true‐population and misspecified models; that is, models with misspecified factor covariance(s) and models with misspecified factor loading(s). The results suggest that, for the ML method, a cutoff value close to .95 for TLI, BL89, CFI, RNI, and Gamma Hat; a cutoff value close to .90 for Mc; a cutoff value close to .08 for SRMR; and a cutoff value close to .06 for RMSEA are needed before we can conclude that there is a relatively good fit between the hypothesized model and the observed data. Furthermore, the 2‐index presentation strategy is required to reject reasonable proportions of various types of true‐population and misspecified models. Finally, using the proposed cutoff criteria, the ML‐based TLI, Mc, and RMSEA tend to overreject true‐population models at small sample size and thus are less preferable when sample size is small.

This study evaluated the sensitivity of maximum likelihood (ML)- generalized least squares (GLS) - and asymptotic distribution-free (ADF)-based fit indices to model misspecification under conditions that varied sample size and distribution. The effect of violating assumptions of asymptotic robustness theory also was examined. Standardized root-mean-square residual (SRMR) was the most sensitive index to models with misspecified factor covariance(s) and Tucker–Lewis Index (1973; TLI)Bollen's fit index (1989; BL89) relative noncentrality index (RNI) comparative fit index (CFI) and the ML- and GLS-based gamma hat McDonald's centrality index (1989; Mc) and root-mean-square error of approximation (RMSEA) were the most sensitive indices to models with misspecified factor loadings. With ML and GLS methods we recommend the use of SRMR supplemented by TLI BL89 RNI CFI gamma hat Mc or RMSEA (TLI Mc and RMSEA are less preferable at small sample sizes). With the ADF method we recommend the use of SRMR supplemented by TLI BL89 RNI or CFI. Finally most of the ML-based fit indices outperformed those obtained from GLS and ADF and are preferable for evaluating model fit.

This study evaluated the sensitivity of maximum likelihood (ML)-, generalized least squares (GLS)-, and asymptotic distribution-free (ADF)-based fit indices to model misspecification, under conditions that varied sample size and distribution. The effect of violating assumptions of asymptotic robustness theory also was examined. Standardized root-mean-square residual (SRMR) was the most sensitive index to models with misspecified factor covariance(s), and Tucker-Lewis Index (1973; TLI), Bollen's fit index (1989; BL89), relative noncentrality index (RNI), comparative fit index (CFI), and the ML- and GLS-based gamma hat, McDonald's centrality index (1989; Mc), and root-mean-square error of approximation (RMSEA) were the most sensitive indices to models with misspecified factor loadings. With ML and GLS methods, we recommend the use of SRMR, supplemented by TLI, BL89, RNI, CFI, gamma hat, Mc, or RMSEA (TLI, Mc, and RMSEA are less preferable at small sample sizes). With the ADF method, we recommend the use of SRMR, supplemented by TLI, BL89, RNI, or CH. Finally, most of the ML-based fit indices outperformed those obtained from GLS and ADF and are preferable for evaluating model fit. (PsycINFO Database Record (c) 2012 APA, all rights reserved)

We tested whether linear structural models of the mechanisms underlying flicker sensitivity could reproduce the variance-covariance matrix of temporal contrast sensitivity data. Monocular sensitivities to frequencies between 2.5 and 45 Hz were measured for 124 subjects, ages 18-88 yr. Exploratory factor analyses revealed that both a two-mechanism and a three-mechanism model could adequately account for the data. Furthermore, confirmatory factor analyses and full structural equation models, using age as an explanatory variable, supported both models, with the three-factor model giving a somewhat better representation of the data. Parsimony favors the two-mechanism model. But patterns of loss associated with pre-exudative age-related maculopathy are more easily understood in terms of three underlying mechanisms.

consider the following issues: (a) the usefulness of the χ[superscript]2 statistic based on various estimation methods for model evaluation and selection; (b) the conceptual elaboration of and selection criteria for fit indexes; and (c) identifying some crucial factors that will affect the magnitude of χ[superscript]2 statistics and fit indexes / review previous research findings as well as report results of some new, unpublished research (PsycINFO Database Record (c) 2012 APA, all rights reserved)

This study examined the relationship of four aspects of psychiatric treatment (use of medication, client-therapist ethnic match, treatment in an Asian-specific clinic, and professional therapist) to participation in treatment and outcome of treatment in low income Asian-American clients (n = 273) of the Los Angeles County mental health system who were diagnosed with major depression. Based on cultural responsiveness theory, the study tested the hypothesis that use of medication in treatment would have the greatest effect on participation and outcome followed, in order, by client-therapist ethnic match, treatment in an Asian-specific clinic, and treatment by a professional therapist. The hypotheses were largely supported: treatment with medication had a significant relationship to total number of treatment sessions (participation) and improvement in the admission-discharge Global Assessment Scale (GAS) score (outcome). Treatment by a therapist of the same ethnicity as the client and treatment in an agency designated to provide services to Asian clients both had significant relationships to the number of treatment sessions but not to GAS score improvement. Four covariates included in the analysis and treatment by a professional therapist had no relationship to either of the dependent variables.

Use of public outpatient mental health services and treatment outcomes were studied among Chinese, Japanese, Filipino, Korean, and Southeast-Asian Americans in Los Angeles County. Filipinos were underrepresented in the system, whereas Southeast Asians were overrepresented and had higher utilization rates, but showed less improvement, than did the other groups. The influence of therapist-client ethnic match and of clinicians' professional status were assessed, and recommendations are made for further research based on present findings.

Covariance structure analysis uses chi 2 goodness-of-fit test statistics whose adequacy is not known. Scientific conclusions based on models may be distorted when researchers violate sample size, variate independence, and distributional assumptions. The behavior of 6 test statistics is evaluated with a Monte Carlo confirmatory factor analysis study. The tests performed dramatically differently under 7 distributional conditions at 6 sample sizes. Two normal-theory tests worked well under some conditions but completely broke down under other conditions. A test that permits homogeneous nonzero kurtoses performed variably. A test that permits heterogeneous marginal kurtoses performed better. A distribution-free test performed spectacularly badly in all conditions at all but the largest sample sizes. The Satorra-Bentler scaled test statistic performed best overall.

Covariance structure analysis uses χ–2 goodness-of-fit test statistics whose adequacy is not known. Scientific conclusions based on models may be distorted when researchers violate sample size, variate independence, and distributional assumptions. The behavior of 6 test statistics was evaluated with a Monte Carlo confirmatory factor analysis study. The tests performed dramatically differently under 7 distributional conditions at 6 sample sizes. Two normal-theory tests worked well under some conditions but completely broke down under other conditions. A test that permits homogeneous nonzero kurtoses performed variably. A test that permits heterogeneous marginal kurtoses performed better. A distribution-free test performed spectacularly badly in all conditions at all but the largest sample sizes. The Santorra-Bentler scaled test performed best overall. (PsycINFO Database Record (c) 2012 APA, all rights reserved)