Geng Academic Fraud Detection Report: Wang Enxu et al. — Study on the Influence Mechanism of Built Environment on Urban Vitality Spatial Differentiation at Street Scale Based on MGWR (Shenyang Central Urban Area)

Academic fraud report · Geng Detector

Summary

This report assesses Wang Enxu, Zhou Jiang, Yang Jun, Wang Yicheng, Yang Pinyan, and Wang Xuejing's article published in Geographical Science (2024; DOI: 10.13249/j.cnki.sgs.20230589). Verdict: highly suspicious. The principal concern is a methodological inconsistency in the construction of the composite 'urban vitality' index in Section 1.3.1 (p. 1324). The authors state that vitality is the mean of spatial vitality, economic vitality, and cultural vitality, but the constituent indicators — a ratio of heat-mean to area, nighttime-light pixel counts, and POI density — are expressed in incommensurate physical units. No standardization (e.g., min-max, z-score) is reported prior to averaging, which would render the composite mathematically ill-defined and dominated by the largest-magnitude component. A secondary issue concerns incomplete statistical reporting: OLS results in Table 2 omit standard errors, t-values, and R²/Adjusted R², while MGWR goodness-of-fit values are cited narratively (e.g., 81.4% explained variance) without the underlying residual statistics. Image-based checks were not feasible from text-only extraction. The report cannot confirm intent; findings should be treated as methodological red flags requiring author response.

Verdict

🟠 Highly suspicious — primarily due to a likely dimensional inconsistency in the composite index construction, compounded by incomplete statistical reporting for the regression models. Image-level forensic checks were not possible with the available material.

Key findings

Dimensional mixing in the urban vitality index (Section 1.3.1, p. 1324): Spatial vitality is defined as a ratio (mean heat / area), economic vitality as a count of nighttime-light pixels, and cultural vitality as POI density. The paper states that urban vitality is computed as the arithmetic mean of these three indicators, yet no standardization (min-max, z-score, range normalization) is mentioned. Averaging quantities in incommensurate units is statistically invalid; the resulting index will be dominated by whichever component has the largest absolute magnitude (most likely POI density or nighttime-light pixel counts).
Incomplete OLS reporting (Table 2, pp. 1325–1326): Several coefficients (population density, bus stops, metro) are reported without significance markers, and the table does not include standard errors, t-values, F-statistic, R², or Adjusted R². The lack of these standard goodness-of-fit and inferential statistics undermines reproducibility.
Incomplete MGWR reporting (§2.3.1): The paper states that MGWR's Adjusted R² exceeds GWR's and that the model can explain 81.4% of variance, but does not provide the explicit MGWR Adjusted R² value, residual sum of squares, or diagnostic statistics needed to evaluate the claim.
Image analysis not performed: Figures 1–3 are referenced as spatial distribution maps. Pixel-level duplication, splicing, and color-band manipulation could not be assessed from text-only extraction; such checks are recommended if source images become available.

Evidence highlights

Composite index formula (Section 1.3.1, p. 1324): "城市活力取城市空间活力、城市经济活力与城市文化活力平均值而得" — three quantities of different physical dimensions averaged without documented normalization.
Table 2 (pp. 1325–1326): OLS regression coefficients listed without SE / t / F / R² / Adjusted R².
§2.3.1: narrative claim of 81.4% explained variance for MGWR without supporting residual diagnostics.
DOI: 10.13249/j.cnki.sgs.20230589

Notes

The dimensional-mixing issue is the most substantive finding and is independently verifiable from the text. It does not, by itself, prove fabrication, but it indicates either a methodological error that invalidates downstream MGWR results or an oversight that must be clarified by the authors.
Selective or incomplete reporting of model statistics is a recognized risk factor for questionable research practice; it is not, on its own, definitive evidence of fraud.
The assessment is limited by text-only access; figure-level forensic analysis and re-analysis of the authors' raw data are required to upgrade or downgrade the verdict.
Recommended follow-up: request the original computation code/data for the three vitality sub-indices and the full MGWR diagnostic output from the corresponding author; if unsatisfied, raise the concerns with the journal editorial office (Geographical Science).