Effect of Sample Size on Goodness-Fit of-Fit Indices in Structural Equation Models

Yadama, Gautam N.; Pandey, Shanta

doi:10.1300/j079v20n03_03

Cited by 65 publications

(50 citation statements)

References 0 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…Among the absolute indicators, the  2 /df value is frequently used as a primary measure of fit, which should be close to unity for correct models, although Wheaton, Muthén, Alwin, and Summers (1977) considered that values up to three are acceptable. However, this index is known to be adversely affected by large samples (Yadama & Pandey, 1995) as in the present study. When the analysis of the final model was repeated on a smaller, randomly chosen sample of 200 participants, the  2 /df value fell to 2.42, which is within the range considered to indicate an adequate correspondence between the model and data.…”

Section: Confirmatory Factor Analysismentioning

confidence: 61%

The Oswald Clergy Burnout Scale: Reliability, Factor Structure and Preliminary Application Among Australian Clergy

Francis

Hills

Kaldor³

2008

Pastoral Psychol

View full text Add to dashboard Cite

This study assesses the Oswald Clergy Burnout Scale (OCBI), the psychometric properties of which have not been previously described. Analysis of responses from a large number (N = 3,012) of ministers in charge of Australian congregations showed that the scale's internal reliability was satisfactory, and that the scale could be represented by two factors, identified as the personal and social aspects of burnout respectively. This structure was supported by confirmatory factor analysis. Several demographic and job-related variables that might relate to burnout were regressed on the total, personal and social factor scores. Age is the predominant (negative) predictor of burnout as measured by the total scale and the personal factor scores. All variables predict burnout as measured by the social factor.However, in all models, the predictor variables account for no more than 5% of the total variance. These findings suggest that demographic factors and working conditions are poor predictors of burnout among clergy.

show abstract

Section: Confirmatory Factor Analysismentioning

confidence: 61%

The Oswald Clergy Burnout Scale: Reliability, Factor Structure and Preliminary Application Among Australian Clergy

Francis

Hills

Kaldor³

2008

Pastoral Psychol

View full text Add to dashboard Cite

show abstract

“…Because therapists only rated their own sessions, interrater reliability was not determined. A confirmatory factor analysis using structural models with AMOS (6.0) software (Arbuckle, 2005) with maximum likelihood estimation showed the two-factor model demonstrated a good fit to the data, namely, a Root Square Mean Square Error of Approximation (RMSEA) of .05 or less (Browne & Cudeck, 1993) and the Normed Fit Index (NFI), Incremental Fit Index (IFI), and Comparative Fit Index (CFI) of .90 or greater (Hu and Bentler, 1995; Kline, 1998; Marsh, Balla, & McDonald, 1988; Yadama & Pandey, 1995). The respective fundamental and advanced strategies fit indices were as follows: RMSEA = .06 and .00 ; NFI = .98 and 1.0; IFI = .99 and 1.0; and CFI = .99 and 1.0.…”

Section: Methodsmentioning

confidence: 99%

Correspondence of motivational enhancement treatment integrity ratings among therapists, supervisors, and observers

Martino¹,

Ball²,

Nich

et al. 2009

Psychotherapy Research

154

106

View full text Add to dashboard Cite

This study examined the correspondence of treatment integrity ratings (adherence and competence) among community program therapists, supervisors, and observers for therapists who used motivational enhancement therapy (MET) within a National Institute on Drug Abuse Clinical Trial Network protocol. The results suggested there was reasonable agreement between the three groups of raters about the presence or absence of several fundamental MET strategies. Moreover, relative to observers, therapists and supervisors were more positive in their evaluations of the therapists' MET adherence and competence. These findings underscore the need for objective monitoring of therapists' performance when using empirically supported treatments and for adequately training therapists and supervisors to evaluate their treatment implementation in community programs, and are consistent with observations that different perspectives on the therapeutic process are not interchangeable.

show abstract

“…All factors were allowed to correlate, and no correlated errors were included in the estimation models. In order to evaluate the fit of the models, observed model covariances were compared with the null hypothesis model (Yadama and Pandey 1995). Fit of any model was assessed by a non-significant χ 2 , incremental fit index (IFI; Bollen 1989) ≥0.90, normalized fit index (NFI; Bentler and Bonett 1980;Marsh et al 1988) ≥0.80, non-normalized fit index (NNFI; Bentler and Bonett 1980) ≥0.90, comparative fit index (CFI; Bentler 1990) ≥0.90, standardized root mean square of errors <0.10 (SRMR; Marsh et al 1988), and root mean square error approximation (RMSEA ;Steiger 1990;Bentler and Bonett 1980;Marsh et al 1988) <0.10.…”

Section: Discussionmentioning

confidence: 99%

An investigation of the validity and reliability of the adapted mathematics anxiety rating scale-short version (MARS-SV) among Turkish students

Baloğlu

2010

Eur J Psychol Educ

View full text Add to dashboard Cite

This study adapted the Mathematics Anxiety Rating Scale-Short Version (MARS-SV) into Turkish and investigated the validity and reliability of the adapted instrument. Twenty-five bilingual experts agreed on the language validity, and 49 Turkish language experts agreed on the conformity and understandability of the scale's items. Thirty-two subject matter experts' responses provided evidence for content validity. Finally, 475 college students (51.58% men) responded to the adapted scale. Based on the existing literature, several models were specified and estimated. Confirmatory factor analyses showed that only five-factor model provided fit to the data (all fit indices exceeded 0.90). Standardized factor loadings were found all positive, significant, and ranged from 0.40 to 0.78. Factors are named as the Mathematics Test Anxiety, Course Anxiety, Computation Anxiety, Application Anxiety, and Social Anxiety. All subscales were significantly correlated with the total scale score as well as among themselves. It was concluded that the adapted scale measures the construct of mathematics anxiety in Turkish college student populations.Résumé Dans cette étude, nous avons adapté l'Echelle de la Graduation de l'Anxiété de la Mathématique, (la forme courte-MARS-SV) en turc et nous avons étudié les caractéris-tiques de la validité et la fiabilité de l'instrument adapté. 25 experts bilingues sont tombés d'accord sur la validité de la langue et 49 experts de La langue Turque se sont mis d'accord sur la conformité et la compréhensibilité des éléments de l'échelle. 32 experts en la matière ont exprimé leurs opinions sur la validité de son contenu. Finalement, 475 étudiants (dont 51,58% sont des hommes) ont répondu à l'échelle adaptée. En nous basant sur la littérature de ce domaine, nous avons estimé et testé quelques modèles. Lés résultats d'analyse factorielle confirmatoire a montré que seule le modèle à cinq dimensions est en conformité avec les donnés de l'analyse (Indices de concordance ont dépassé 90). Dans ce modèle les valeurs standardisées de facteur sont tous positives et significatives, elles varient de 40 à 78.

show abstract

Effect of Sample Size on Goodness-Fit of-Fit Indices in Structural Equation Models

Cited by 65 publications

References 0 publications

The Oswald Clergy Burnout Scale: Reliability, Factor Structure and Preliminary Application Among Australian Clergy

The Oswald Clergy Burnout Scale: Reliability, Factor Structure and Preliminary Application Among Australian Clergy

Correspondence of motivational enhancement treatment integrity ratings among therapists, supervisors, and observers

An investigation of the validity and reliability of the adapted mathematics anxiety rating scale-short version (MARS-SV) among Turkish students

Contact Info

Product

Resources

About