Feipeng Li

Prof. Feipeng Li


Research Focus & Metrics

Academic Career Length

1Year (2026-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
3 November 2026 in Preprints.org
Verified
Preprint
Open access
PDF file
JWJunqiang Wu
Junqiang Wu
FLFeipeng Li
Feipeng Li
WHWenfeng He
Wenfeng He
XWXibo Wu
Xibo Wu
Ratio-type measures of structural association compare observed co-occurrence with what the marginal distribution would produce, and they return a number whether or not that number can be read. EduCEQ-Toolkit is a Python framework that declines to do so: on a public panel of 4,785 U.S. higher-education institutions, its runtime contract reports that no bandwidth between 50 and 500 km is decidable, and the call returns a structured non-decidability object instead of a coefficient. The framework generalizes the co-location quotient to an arbitrary non-negative pairwise weight, so distance, attribute similarity and network proximity are handled by one estimator. It implements three composable kernel families, an exact first- and second-order decomposition, conditional permutation inference with false discovery rate control, and a three-step contract binding the population count, the neighbourhood count and the coverage proportion. We present the contract as a reusable design pattern for research software: where a condition on a method's validity can be computed, it belongs in the execution path rather than in the manual. Evaluation covers threshold calibration, a decidable case, a refusal case on public data, and computational cost.

10 November 2026 in Preprints.org
Verified
Preprint
Open access
PDF file
JWJunqiang Wu
Junqiang Wu
FLFeipeng Li
Feipeng Li
WHWenfeng He
Wenfeng He
XWXibo Wu
Xibo Wu
Ratio-type measures of spatial association, such as the co-location quotient, depart from unity for two incompatible reasons: interaction between categories, or an uneven distribution that mechanically forces co-occurrence away from the global expectation. We separate the two on a thirty-wave panel of U.S. higher-education institutions (1996–2025; 4,785 institutions in cross-section, 174,479 institution–wave records), decomposing the quotient exactly into a first-order term built from local type shares and a second-order term defined as their ratio, with independent placement rather than global randomization as the null. Across two classification dimensions, three bandwidths and all thirty waves, co-location values span 0.787–1.294, first-order terms reproduce them to within 0.013, and second-order terms remain within 0.980–1.008, widening to 0.971–1.008 at 100 km; the between-wave standard deviation never exceeds 0.0013, an order of magnitude below the deviation it describes. The 2008 financial crisis, a 2012 eligibility rule and the 2020 pandemic leave no trace larger than 0.0012 in the second-order terms, and a spatial difference-in-differences design built around the 2012 rule fails its own placebo and pre-trend diagnostics. Structural indicators should be decomposed before interpretation, and scale change should be reported separately from structural change.