Research on a Bayesian cohort model including explanatory variables other than age, period, and cohort effects
| Project/Area Number |
10680324
|
|---|
| Research Category |
Grant-in-Aid for Scientific Research (C)
|
| Allocation Type | Single-year Grants |
|---|
| Section | 一般 |
|---|
| Research Field |
Statistical science
|
|---|
| Research Institution | The Institute of Statistical Mathematics |
|---|
Principal Investigator |
NAKAMURA Takashi Department of Statistical Methodology, The Institute of Statistical Mathematics, Professor, 調査実験解析研究系, 教授 (20132699)
|
|---|
| Co-Investigator(Kenkyū-buntansha) |
MAEDA Tadahiko Department of Statistical Methodology, The Institute of Statistical Mathematics, Assistant Professor, 調査実験解析研究系, 助手 (10247257)
SAKAMOTO Yoshiyuki Department of Statistical Methodology, The Institute of Statistical Mathematics, Professor, 調査実験解析研究系, 教授 (50000211)
|
|---|
| Project Period (FY) |
1998 – 1999
|
| Project Status |
Completed (Fiscal Year 1999)
|
| Budget Amount *help |
¥2,300,000 (Direct Cost: ¥2,300,000)
Fiscal Year 1999: ¥800,000 (Direct Cost: ¥800,000)
Fiscal Year 1998: ¥1,500,000 (Direct Cost: ¥1,500,000)
|
|---|
| Keywords | age effects / period effects / cohort effects / identification problem / Bayesian model / cohort analysis / Japanese national character / time-series social survey data / 年齢別回答分布 |
|---|
| Research Abstract |
Cohort analysis is a method of separating the effects of age, historical period, and birth time (cohort) from time-series social survey data classified by age group and survey period. Although the method is useful for understanding the mechanism of social change, it is known that the method confronts the identification problem that the three kinds of effects cannot be separated without some prior information. In order to overcome the problem, Nakamura proposed a Bayesian cohort model with the gradually-changing-parameter assumption and a model scheme using Akaike's Bayesian information criterion ABIC.
The purpose of the present research is to develop a new extended cohort model including additional explanatory variables other than age, period, and cohort effects. Examples of these variables are the trend of business conditions, the indicator of different survey system, the cohort size, and so on.
The design matrix of the proposed model is set up and ABIC is derived. By analyzing artificial datasets, the performance of the model is examined. Actual datasets such as household saving rates, crime rates, and percentages concerning religious attitude are analyzed using the models with including the trend of business conditions, the cohort size, and the indicator of different survey system, respectively.
|
Report
(3 results)
Research Products
(4 results)
