When the model specification was held fixed, estimator choice did not change the substantive conclusion for this two-item measure: weight-related anxiety and self-reported weight-management goal non-attainment among undergraduates at a public univers

Background Weight-related anxiety has been widely studied as a correlate of weight-management goal non-attainment among university students, yet it remains unclear whether brief, floor-skewed anxiety measures yield estimator-dependent conclusions when maximum likelihood (ML) and weighted least squares mean- and variance-adjusted (WLSMV) estimation are compared. Holding the model specification constant, this study estimated the anxiety–outcome association and tested whether the two estimators reached the same conclusion. Methods We analyzed cross-sectional survey data from 754 undergraduates (71.4% female; mean age 20.66 years) enrolled at a public university in Southwestern China. Multiple logistic regression was used to identify factors associated with goal non-attainment. A nine-path manifest-variable recursive path model was then estimated under both ML and WLSMV using an identical specification, and direct and indirect associations were obtained from 5,000 bootstrap resamples (percentile 95% CIs), with all coefficients standardized. Results The two-item anxiety composite (range 1–6) showed a marked floor effect, with 39.4% of participants scoring at or below 2.0. In the logistic regression, weight-related anxiety was the strongest correlate of non-attainment (OR = 2.11, 95% CI 1.79–2.49, p < 0.001). In the path model, the standardized anxiety–outcome coefficient was β = 0.436 under ML and β = 0.522 under WLSMV; both estimates were significant and directionally consistent. Eight of nine paths agreed in direction, the single exception being the disordered-eating → outcome path, which was non-significant under both estimators. Under ML, the single-equation latent-response coefficient was identical to the path coefficient ( β = 0.436); under WLSMV, the corresponding values were 0.501 (single-equation) and 0.522 (path). The indirect association was β = 0.119 (95% CI 0.084–0.157), whereas the direct association was non-significant. Conclusion When the specification was held constant, ML and WLSMV yielded the same substantive conclusion, although measurement sufficiency was not established. When the specification varied, WLSMV was more sensitive to path specification than ML was, so the estimator and specification effects could not be disentangled. Weight-related anxiety was associated with goal non-attainment, but the cross-sectional design precluded causal inference.

Authors

Institutions

Publication Details

Journal
Frontiers in Nutrition
Published
2026-11-16
DOI
https://doi.org/10.3389/fnut.2026.1960890
Primary Topic
Psychometric Methodologies and Testing
Type
article
Field-Weighted Citation Impact
0.00
Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
NOV
article

When the model specification was held fixed, estimator choice did not change the substantive conclusion for this two-item measure: weight-related anxiety and self-reported weight-management goal non-attainment among undergraduates at a public univers

Huimin Zhou, Xiao Huang, Huifang Chen
Frontiers in Nutrition
Psychometric Methodologies and Testing
article

When the model specification was held fixed, estimator choice did not change the substantive conclusion for this two-item measure: weight-related anxiety and self-reported weight-management goal non-attainment among undergraduates at a public univers

Huimin Zhou, Xiao Huang, Huifang Chen
article en

Abstract

Background Weight-related anxiety has been widely studied as a correlate of weight-management goal non-attainment among university students, yet it remains unclear whether brief, floor-skewed anxiety measures yield estimator-dependent conclusions when maximum likelihood (ML) and weighted least squares mean- and variance-adjusted (WLSMV) estimation are compared. Holding the model specification constant, this study estimated the anxiety–outcome association and tested whether the two estimators reached the same conclusion. Methods We analyzed cross-sectional survey data from 754 undergraduates (71.4% female; mean age 20.66 years) enrolled at a public university in Southwestern China. Multiple logistic regression was used to identify factors associated with goal non-attainment. A nine-path manifest-variable recursive path model was then estimated under both ML and WLSMV using an identical specification, and direct and indirect associations were obtained from 5,000 bootstrap resamples (percentile 95% CIs), with all coefficients standardized. Results The two-item anxiety composite (range 1–6) showed a marked floor effect, with 39.4% of participants scoring at or below 2.0. In the logistic regression, weight-related anxiety was the strongest correlate of non-attainment (OR = 2.11, 95% CI 1.79–2.49, p < 0.001). In the path model, the standardized anxiety–outcome coefficient was β = 0.436 under ML and β = 0.522 under WLSMV; both estimates were significant and directionally consistent. Eight of nine paths agreed in direction, the single exception being the disordered-eating → outcome path, which was non-significant under both estimators. Under ML, the single-equation latent-response coefficient was identical to the path coefficient ( β = 0.436); under WLSMV, the corresponding values were 0.501 (single-equation) and 0.522 (path). The indirect association was β = 0.119 (95% CI 0.084–0.157), whereas the direct association was non-significant. Conclusion When the specification was held constant, ML and WLSMV yielded the same substantive conclusion, although measurement sufficiency was not established. When the specification varied, WLSMV was more sensitive to path specification than ML was, so the estimator and specification effects could not be disentangled. Weight-related anxiety was associated with goal non-attainment, but the cross-sectional design precluded causal inference.

Frontiers in NutritionVol. 13
Peking University (CN), Beijing Jishuitan Hospital (CN)
Good health and well-being
Psychometric Methodologies and Testing
AI Navigator

Ask Laika to Summarize, Analyze, and Connect papers live on the map.

Summarize Papers & Methodologies

Extract key findings, datasets, and comparative methods across publications.

Benchmark Rankings & Visual Analytics

Rank top research institutions, authors, funders, topics, and journals by Field-Weighted Citation Impact (FWCI) and paper volume with instant charts.

Connect Distant Disciplines

Bridge topological clusters on the map to find hidden collaborative intersections.

Laika access coming soon