Temporal aggregation bias and Gerrymandering urban time series

Stehle, Samuel

doi:10.1007/s10707-021-00452-z

Cited by 1 publication

(2 citation statements)

References 35 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Both GWR and ECSRE show completely different relationships between the predictors and the responses. And this confirms other findings and conclusions found in the literature by Stehle (2022), Wade et al (2019), and others.…”

Section: Resultssupporting

confidence: 93%

“…The former is the effect of spatial aggregation and how highfrequency variables bias the estimation results (Colin et al, 2018); the latter is the distribution abnormalities (e.g., spatial clustering) of confounding variables (Kleinschmidt et al, 2016;Davidson, 2009). Many research results indicate that by increasing data resolution (i.e., definition), such as finer-scale urban data, ecological errors would soar (Stehle, 2022;Wade et al, 2019;Vlahogianni and Karlaftis, 2009).…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

A model for eliminating aggregate and specification bias in multivariate fine-scale urban scenarios

Salehi,

Beni,

Halabian

et al. 2024

Preprint

View full text Add to dashboard Cite

The spatial relationships between predictors and responses are influenced by their frequency and spatial distribution. Ecological bias in regression models can occur due to the aggregate frequency and clustering of independent variables, leading to false, over-, or underestimations. This can be exacerbated by an increase in data resolution, complexity, and variable count, as is often the case in urban research scenarios. To address this issue, a new relationship-estimation model called the Ecologically Corrected Spatial Relationship Estimator (ECSRE) was proposed and compared to Geographically Weighted Regression (GWR). The results showed that ECSRE outperformed GWR by correctly revealing pre-planned relationships in simulated data, presenting a lower influence of aggregate frequencies on the outcome, better suppression of specification errors, higher R2 scores, and better randomness of residuals.

show abstract

Section: Resultssupporting

confidence: 93%