Series estimation under cross-sectional dependence

Nonparametric specification testing via the trinity of tests

“…Lee and Robinson (2013) consider series estimation in the presence of cross-sectional dependence, and we can easily accommodate this with (4.12) amended to the correct rates. Indeed, following that paper let x i be identically distributed with pdf f (x) and the joint pdf of of x i , x j be f ij (x, y), x, y ∈ X.…”

“…Introduce the bivariate dependence measure 13) and replace (4.12) with 14) to obtain tests with Property C, assuming Υ n affords a non-null choice set of γ. The latter is obtained from Assumption B.3 in Lee and Robinson (2013), noting that we have taken iid disturbances as opposed to their more general linear process specification. Indeed specification testing can undoubtedly extend to models such as the SAR in (2.11), where the linear regression component is replaced by a nonparametric function which is then estimated by series and tested for linearity as in the previous paragraph.…”

Inference on modelling cross-sectional dependence for a varying-coefficient model

“…In order to impose the crosssectional dependence, we follow Robinson (2011) and Lee and E-mail address: Bin.Peng@uts.edu.au. Robinson (2016) and denote that…”

“…Pesaran, 2006, Bai, 2009 andso forth). Recently, Robinson (2011) and Lee and Robinson (2016) have employed the time series technique to model the dependence among cross-sectional data sets. Following the spirit of their work, I consider a varyingcoefficient model with cross-sectional dependence in this study.…”

Optimal Uniform Convergence Rates and Asymptotic Normality for Series Estimators Under Weak Dependence and Weak Conditions

“…data and Lee and Robinson (2013) for spatially dependent data. But the uniform convergence rates obtained in these papers are slower than the optimal rate of Stone (1982).…”