Sensitivity of Goodness of Fit Indexes to Lack of Measurement Invariance

By F. Chen. 2007

11,070 citations recorded by the source corpus.

Subfields: Information Systems and Management, Management Science and Operations Research

View Paper: https://www.semanticscholar.org/paper/246e9cb29b534cb4e120a38d5039e1d2a858ec51

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Information Systems and Management

In Management Science and Operations Research

Citations in the corpus

Development and initial validation of the Life Skills Scale for Sport (2017, methodology): “When comparing models, similar fit is evident when changes in the CFI are < 0.01 and increases in RMSEA are < 0.015 (Chen, 2007; Cheung & Rensvold, 2002).”
Measuring Teachers’ enjoyment, anger, and anxiety: The Teacher Emotions Scales (TES) (2016, methodology): “We inspected measurement invariance of the TES across the Canadian sample (this study) and the German sample (Study 1) by applying a commonly used stepwise procedure (e.g., Chen, 2007; Van de Schoot, Lugtig, & Hox, 2012).”
Achievement Emotions and Academic Performance: Longitudinal Models of Reciprocal Effects. (2017, methodology): “To establish equivalence of constructs for analyzing correlations and path coefficients, metric invariance is the minimum needed (Chen, 2007; Steenkamp & Baumgartner, 1998).”
University Instructors’ Achievement Goals for Teaching (2019, methodology): “Differences in model fit between subsequent models were evaluated using 2 LL rescaled difference in the model log-likelihood values (robust chi-square difference test), as well as the differences in CFI and RMSEA (using the commonly suggested cut-off values of CFI .02 and RMSEA .015; Chen, 2007).”
Assessing the Hypothesis of Measurement Invariance in the Context of Large-Scale International Surveys (2014, methodology): “TALIS researchers supported their findings regarding measurement invariance with research that was limited in scope to few groups and relatively small sample sizes (Chen, 2007; French & Finch, 2006).”
Testing Measurement Invariance and Comparing Latent Factor Means Within a Confirmatory Factor Analysis Framework (2011, methodology): “However, their study did not explore various types of noninvariance (uniform or mixed noninvariance), which Chen (2007) indicated significantly influences the model fit results.”
Academic buoyancy and psychological risk: Exploring reciprocal relationships ☆ (2013, positive): “…two significant cross-lagged parameters (i.e., all models except failure avoidance), there was no significant difference in model fit (based on CFI criteria recommended by Cheung & Rensvold, 2002 and on RMSEA criteria recommended by Chen, 2007) when the cross-lagged paths were fixed to be equal.”
When passion fades: disentangling the temporal dynamics of entrepreneurial passion for founding (2016, methodology): “Using commonly applied statistical cut-off points for key fit measures (i.e., CFI, RMSEA, SRMR; see Chen, 2007), the scalar invariance model was found to be adequate for all individual multi-item constructs we measured over time (i.e., role ambiguity and intense positive feelings).”
Measuring Socioeconomic Background in PISA: One Size Might not Fit all (2013, methodology): “Using Chen (2007) as a general guideline, we viewed changes in CFI > –.01 and changes in RMSEA .01 as evidence of an acceptable level of invariance.”
Methodological Measurement Fruitfulness of Exploratory Structural Equation Modeling (ESEM): New Approaches to Key Substantive Issues in Motivation and Engagement (2011, background): “Cheung and Rensvold (2001) and Chen (2007) suggested that if the decrease in fit for the more parsimonious model is less than 0.01 for incrementalfit indices like the CFI, then there is reasonable support for the more parsimonious model.”