Effect of Spatial Data Units on the Variance of a Regression Model Containing a Distance Variable Between a Pre-determined Point and the Spatial Data Unit
-
- TAGASHIRA Naoto
- Central Research Institute of Electric Power Industry
Bibliographic Information
- Other Title
-
- 地区単位で集計された距離を説明変数とする回帰モデルの安定性
- チク タンイ デ シュウケイサレタ キョリ オ セツメイ ヘンスウ ト スル
- リュウシュツ モデル ノ コウチク ニ オケル GIS ノ オウヨウ ニカンス
Search this article
Abstract
This paper investigates a problem which arises when we wish to estimate a regression coefficient using a distance variable between a pre-determined point and a spatial data unit. First, we show that the variance of the regression coefficient decreases with the number of the units and propose a desirable number of the units whose variance is close to that of a disaggregated model. Second, we obtain the shapes of the units that achieve the minimum and the maximum variances assuming a constant number of the units to examine the influence of the shape of the units on the variance.
Journal
-
- Theory and Applications of GIS
-
Theory and Applications of GIS 6 (1), 29-38, 1998
Geographic Information Systems Association
- Tweet
Details 詳細情報について
-
- CRID
- 1390282679451601280
-
- NII Article ID
- 10020375596
- 10010540527
- 10020375589
-
- NII Book ID
- AN10457305
-
- ISSN
- 21855633
- 13405381
-
- Data Source
-
- JaLC
- NDL
- Crossref
- CiNii Articles
-
- Abstract License Flag
- Disallowed