Effective sample size for spatial regression models

Acosta, Jonathan; Vallejos, Ronny

doi:10.1214/18-ejs1460

Cited by 16 publications

(7 citation statements)

References 31 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…How to optimise the power of the test in specific settings should be pursued. Next, eliminating the need for generating Monte Carlo samples from the null point process may be possible [Acosta et al, 2018]. We found that 'effective-sample size'-adjusted rank correlation tests showed very poor performance.…”

Section: Discussionmentioning

confidence: 98%

A fast Monte Carlo test for preferential sampling

Watson

2020

Preprint

View full text Add to dashboard Cite

The preferential sampling of locations chosen to observe a spatio-temporal process has been identified as a major problem across multiple fields. Predictions of the process can be severely biased when standard statistical methodologies are applied to preferentially sampled data without adjustment. Currently, methods that can adjust for preferential sampling are rarely implemented in the software packages most popular with researchers. Furthermore, they are technically demanding to design and fit. This paper presents a fast and intuitive Monte Carlo test for detecting preferential sampling. The test can be applied across a wide range of data types. Importantly, the method can also help with the discovery of a set of informative covariates that can sufficiently control for the preferential sampling. The discovery of these covariates can justify continued use of standard methodologies. A thorough simulation study is presented to demonstrate both the power and validity of the test in various data settings. The test is shown to attain high power for non-Gaussian data with sample sizes as low as 50. Finally, two previously-published case studies are revisited and new insights into the nature of the informative sampling are gained. The test can be implemented with the R package PStestR.

show abstract

Section: Discussionmentioning

confidence: 98%

A fast Monte Carlo test for preferential sampling

Watson

2020

Preprint

View full text Add to dashboard Cite

show abstract

“…It seems reasonable to assume that the variance of 𝑝𝐻 is proportional to the sum of the variances of the parameters. Acosta and Vallejos (2018) showed that the information from 𝑁 correlated samples is equivalent to the information matrix for 𝑛 * independent samples. Thus, increasing the number of grids in a spatial field has greatly diminishing returns as the correlation between contiguous grids grows.…”

Section: Discussionmentioning

confidence: 99%

“…An analytical solution of 𝐶 −1 𝜂 is impractical for the exponential covariance function used here. But if we consider an increasing function of cost for the number of grids and the exponential correlation function, we know that the 𝑁𝑡𝑟(𝑀 𝑡 𝐶 −1 𝜂 𝑀) is a decreasing function of the range 𝑎 𝐵𝑝ℎ (Acosta & Vallejos, 2018). Greater spatial autocorrelation means that lesser information is gained by adding more (sample size) and so we can conclude that as 𝑎 𝐵𝑝ℎ gets bigger and bigger, the optimal number of grids will be smaller.…”

Section: Discussionmentioning

confidence: 99%

Optimal grid size for site‐specific nutrient application

Mills,

Brorsen,

Poursina

et al. 2023

Agricultural Economics

View full text Add to dashboard Cite

This article develops a theoretical framework to determine the economically optimal grid size when sampling to guide precision application of inputs. Our theoretical model shows that a finer grid is optimal with increases in output price. A coarser grid is optimal with increased sampling costs and increased spatial correlation. With an exponential variogram, for example, the optimal grid size increases as the range increases. An applied example of lime application for winter wheat is used to demonstrate the framework. Spatial variograms for pH and Buffer pH were estimated using data from nine farmer fields. At the average spatial variability from these fields and using typical Oklahoma wheat yields of 40 bu/ac, grid sampling was economical when prices were above $5.43/bu. The optimal grid sizes were slightly larger than the 2.5‐acre sizes that are typically used by producers. The empirical example verified the theory in that it found that optimal grid size decreased with higher output prices and crop yields and increased with greater spatial correlation. For example, optimal grid size approximately doubled as the range of the exponential correlation function doubled.

show abstract

“…The authors argue that small area estimation methods, "borrow strength" from adjacent regions and therefore compensate for the small sample sizes which is often observed in these areas. In turn, Acosta and Valejos ( 2018 ) provide a simulation of information loss depending on the n in spatial regression. They show that including 25 units provides the evidence for spatial correlation only slightly weaker than including 50 units.…”

Section: Public Data Biasmentioning

confidence: 99%

Spatial matching on the urban labor market: estimates with unique micro data

Woźniak

2021

J Labour Market Res

View full text Add to dashboard Cite

In the paper, we investigate spatial relationship on the labor market of Poznań agglomeration (Poland) with unique data on job vacancies. We have developed spatial panel models to assess the search and matching process with a particular focus on spatial spillovers. In general, spatial models may provide different findings than regular panel models regarding returns to scale in matching technology. Moreover, we have identified global spillover effects as well as other factors that impact the job-worker matching. We underline the role of data on job vacancies: the data retrieved from commercial job portals produced much more reliable estimates than underestimated registered data.

show abstract

Effective sample size for spatial regression models

Cited by 16 publications

References 31 publications

A fast Monte Carlo test for preferential sampling

A fast Monte Carlo test for preferential sampling

Optimal grid size for site‐specific nutrient application

Spatial matching on the urban labor market: estimates with unique micro data

Contact Info

Product

Resources

About