The art and science of econometrics / Ping Zong.

Format:

Book

Author/Creator:

Zong, Ping, author.

Series:

Routledge Studies in Economic Theory, Method and Philosophy

Language:

English

Subjects (All):

Econometrics.

Physical Description:

1 online resource (249 pages)

Place of Publication:

Abingdon, Oxon ; New York, NY : Routledge, [2022]

Summary:

"Today econometrics has been widely applied in the empirical study of economics. As an empirical science, econometrics uses rigorous mathematical and statistical methods for economic problems. Understanding the methodologies of both econometrics and statistics is a crucial departure for econometrics. The primary focus of this book is on providing an understanding of statistical properties behind econometric methods. Following the introduction in Chapter 1, Chapter 2 provides the methodological review of both econometrics and statistics in different periods since the 1930s. Chapters 3 and 4 explain the underlying theoretical methodologies for estimated equations in the simple regression and multiple regression models and discuss the debates about P-values in particular. This part of the book offers the reader a richer understanding of the methods of statistics behind the methodology of econometrics. The latter five chapters of the book are focused on the discussion of regression models using time series data, traditional causal econometric models and the latest statistical techniques. By concentrating on dynamic structural linear models like state-space models and the Bayesian approach, the book alludes to the fact that this methodological study is not only a science but also an art. This work serves as a handy reference book for anyone interested in econometrics, particularly in relevance to students, academic and business researchers in all quantitative analysis fields"-- Provided by publisher.

Contents:

Cover

Half Title

Series Page

Title Page

Copyright Page

Dedication

Contents

List of figures

Preface

1. Introduction

1.1. Introduction

1.2. The motivation for writing this book

1.3. The focus of this book

1.4. The scope and structure of the book

2. Methodology Review

2.1. Introduction

2.2. The creation of econometrics

2.3. Social physics

2.3.1. Deductive and inductive

2.3.2. Statistical induction

2.3.3. Motivations of creating social physics

2.3.4. 'Mathematic games' vs. science for econometrics

2.4. Core ideas of science in econometrics

2.4.1. Error assessment and probability application

2.4.2. Stochastic and deterministic

2.4.3. The 'Error' concept in modern statistics

2.5. Probability revolution in econometrics

2.5.1. Koopmans and Haavelmo's important contributions

2.5.2. Various estimation methods as an approximation

2.5.3. Theoretical assumption tests

2.5.4. Probability approach in time series analysis

2.6. 'Discovering' from data

2.6.1. The econometric model approach

2.6.2. Box-Jenkins' ARIMA approach

2.7. From static to dynamic modelling

2.7.1. Structural econometric time series approach

2.7.2. Dynamic linear models

2.7.3. The Bayesian and non-Bayesian approaches

2.8. Debates of statistical significance in the 21st century

2.8.1. Debates on p-values

2.8.2. The reasons for stopping using p-values

2.8.3. Disagreed opinions

2.9. Summary and conclusions

3. Methodology of Simple Regression Analysis

3.1. Introduction

3.2. Simple regression analysis

3.2.1. Population regression functions (PRF)

3.2.2. Sample regression functions (SRF)

3.2.3. The relationship between PRF and SRF

3.3. Estimation methods for SRF coefficients

3.3.1. Population mean and sample mean

3.3.2. Least square techniques.

3.4. Standard statistical assumptions

3.4.1. Statistical assumptions of the PRF estimation

3.4.2. The normal distribution in the regression model

3.5. Best linear unbiased estimators (BLUE)

3.5.1. The unbiased estimate

3.5.2. Efficiency

3.5.3. The precision of regression coefficients

3.6. Accuracy assessments and model valuations

3.6.1. The residual standard error (RSE)

3.6.2. 'Goodness of fit' - R2 statistic

3.7. The null hypothesis statistical testing

3.7.1. The confidence interval estimation

3.7.2. The statistical significant test

3.8. The crisis of statistical science in p-values

3.8.1. Misinterpretation of the p-values

3.8.2. Misuse of the p-values

3.8.3. Arbitrary threshold value of p &lt

0.05

3.8.4. Practical alternatives to p-values

3.9. The simulation approach

3.10. Summary and conclusions

4. Methodology of Multiple Regression Analysis

4.1. Introduction

4.2. Methodology of multiple linear regressions

4.3. Least square estimation in multiple regression models

4.4. Assumptions of multiple regression models

4.4.1. Assumptions for multiple regression models

4.4.2. The precision of multiple regression coefficients

4.5. Problems with multiple regression models

4.5.1. Specification bias

4.5.2. Detecting the presence of unnecessary variables

4.6. Multicollinearity

4.6.1. The perfect multicollinearity

4.6.2. The imperfect multicollinearity

4.6.3. The sequence of multicollinearity

4.6.4. Detecting problematic multicollinearity

4.6.5. Removing the multicollinearity

4.7. Lasso regression

4.7.1. Over-fitting problems

4.7.2. Lasso regression techniques

4.7.3. Lasso regression equations

4.7.4. Machine learning

4.8. Summary and conclusions

5. Time Series Regression Analysis

5.1. Introduction.

5.2. Challenges of the regression models using time series data

5.2.1. Autocorrelations

5.2.2. Covariance (weak) stationary time series

5.2.3. Autocovariance and autocorrelation functions

5.3. Regression models using time series data

5.3.1. Econometric lagged models

5.3.2. Autocorrelated disturbance models

5.4. Detection of autocorrelation

5.4.1. Durbin-Watson test

5.4.2. Durbin h statistic test

5.4.3. Plotting the graph of the correlogram

5.5. Estimation of the autocorrelation coefficient

5.5.1. The non-iterative approach

5.5.2. The iterative approach

5.6. Non-stationary time series

5.6.1. Random work

5.6.2. 'Stochastic trends' vs. 'deterministic trends'

5.7. Challenges of traditional econometric models

5.7.1. The traditional econometric modelling approach

5.7.2. The time series analysis approach

5.8. Summary and conclusions

6. Methodologies of Economic Forecasting

6.1. Introduction

6.2. Causal econometric forecasting models

6.2.1. Sample regression models

6.2.2. Forecasting errors

6.2.3. Potential sources of error

6.2.4. Multiple regression forecasting models

6.3. Art of ARIMA models

6.3.1. General frameworks of the ARIMA process

6.3.2. AR, MA, ARMA, and ARIMA processes

6.3.3. Determination of ARIMA models

6.3.4. Model estimation and its coefficients

6.3.5. Model evaluation and diagnostic checking

6.3.6. The Box-Jenkins forecasting procedure

6.3.7. Forecasting accuracy

6.4. Summary and conclusions

7. Dynamic Econometric Models

7.1. Introduction

7.2. Simultaneous equation models

7.2.1. General ideas of SEM

7.2.2. Structural and reduced forms

7.3. Error correction models

7.3.1. A specialised case of ARIMA and SEM

7.3.2. The ECM approach

7.3.3. Estimation methods

7.3.4. ECM for the non-stationary series.

7.4. Structural econometric time series approach

7.4.1. The formulation of SETSA

7.4.2. Dynamic structural econometric models

7.5. State space models

7.5.1. State space frameworks

7.5.2. General state space models

7.5.3. Kalman filter

7.5.4. Pseudo error autocorrelation approach

7.5.5. Hyperparameter estimations

7.6. Univariate state space models

7.6.1. The basic structural model (BSM)

7.6.2. USSM estimations

7.7. Time-invariant models (TIM) and time-varying models (TVM)

7.7.1. SSM including regression models

7.7.2. A time-invariant mode settings

7.7.3. A time-varying model settings

7.8. Multivariate state space models

7.8.1. The extension of the USSM

7.8.2. The formulation of MSSM

7.8.3. Initialisations and assumptions

7.8.4. The extended MSSM

7.9. Summary and conclusions

8. The Bayesian Approach

8.1. Introduction

8.2. Differences between frequentists and Bayesians

8.2.1. Two philosophies about probability

8.2.2. Two different probability reasoning

8.2.3. Debates between frequentists and Bayesians

8.3. Learning from experience

8.3.1. Bayesian networks and the conditional probability

8.3.2. The joint probability

8.3.3. The dynamic Bayesian network

8.4. The mechanism of Bayesian inferences

8.4.1. Bayesian inference on the regression coefficients

8.4.2. The Bayesian theorem

8.4.3. Bayesian point parameter estimate

8.4.4. Bayes point prediction

8.5. The mechanism of Bayesian forecasting

8.5.1. Bayesian predict in random walk process

8.5.2. Bayesian predict in a multivariate model

8.5.3. The use of modern simulation techniques

8.6. Advantages of the Bayesian approach

8.7. Summary and conclusions

9. Résumé

9.1. Introduction

9.2. Causality methodology in econometrics

9.3. Core ideas of the econometric methodology.

9.4. Notable development in the last few decades

9.5. Challenges

9.6. Conclusion

Bibliography

Index.

Notes:

Includes bibliographical references and index.

Description based on print version record.

Other Format:

Print version: Zong, Ping The Art and Science of Econometrics

ISBN:

1-00-327390-4

1-003-27390-4

1-000-58022-9

9781003273905

OCLC:

1287744284