STATISTICAL MISSPECIFICATION AND THE RELIABILITY OF INFERENCE: THE SIMPLE T-TEST IN THE PRESENCE OF MARKOV DEPENDENCE
0
조회
0
다운로드
0
인용
0
참고문헌
키워드
초록
The aim of this paper is to consider the problem of unreliable statistical inference caused by the presence of statistical misspecification, and discuss the merits of alternative ways to address the problem like invoking generic robustness results or/and using nonparametric inference. For simplicity the discussion focuses on the t-test for hypotheses concerning the mean in the context of the simple Normal model, with the misspecification coming in the form of Markov Dependence (MD). By deriving explicitly the nominal and actual error probabilities, it is shown that the presence of MD turns the t-test into an unreliable procedure. It is argued that invoking traditional robustness arguments can often be very misleading and, in general, this strategy does not address the unreliability of inference problem, even if one were to use the actual error probabilities. A more appropriate strategy is to respecify the original statistical model to account for the misspecification, and test the hypotheses of interest using an inference procedure that is optimal in the context of the respecified model. It is shown that the presence of MD gives rise to the Autoregression (AR(1)) as the respecified model, and one can test the original hypotheses concerning the mean. The optimal t-test in the context of the AR(1) is shown to be related but different from the original and the modified t-tests.
Abstract
The aim of this paper is to consider the problem of unreliable statistical inference caused by the presence of statistical misspecification, and discuss the merits of alternative ways to address the problem like invoking generic robustness results or/and using nonparametric inference. For simplicity the discussion focuses on the t-test for hypotheses concerning the mean in the context of the simple Normal model, with the misspecification coming in the form of Markov Dependence (MD). By deriving explicitly the nominal and actual error probabilities, it is shown that the presence of MD turns the t-test into an unreliable procedure. It is argued that invoking traditional robustness arguments can often be very misleading and, in general, this strategy does not address the unreliability of inference problem, even if one were to use the actual error probabilities. A more appropriate strategy is to respecify the original statistical model to account for the misspecification, and test the hypotheses of interest using an inference procedure that is optimal in the context of the respecified model. It is shown that the presence of MD gives rise to the Autoregression (AR(1)) as the respecified model, and one can test the original hypotheses concerning the mean. The optimal t-test in the context of the AR(1) is shown to be related but different from the original and the modified t-tests.
참고문헌 (40)
- Anderson,T.W, "The Statistical Analysis of Time Series" (1971), Wiley, NY
- Arnold,S.F, "Mathematical Statistics" (1990), Prentice-Hall, NJ.
- Box,G.E.P, "Non-Normality and Tests on Variance", Biometrika, pp.318~335 (1953)
- Box,G.E.P, "“Robustness in the Strategy of Scientific Model Building,”in Robustness in Statistics, ed. by Launer" (1979), R. L. and G. N. Wilkinson,Academic Press, NY
- Box, G. E. P, "Permutation Theory in the Derivation of Robust Criteria and the Study of Departures from Assumptions,”", Journal of the Royal Statistical Society,, 17, pp.1~34 (1955)
- Box, G. E. P, "Time Series Analysis: Forecasting and Control" (1970), Holden-Day, San Francisco.
- Box, G. E. P, "Bayesian Inference in Statistical Analysis" (1973), Wiley, NY
- Cox,D.R, "Fieller’s Theorem and a Generalization", Biometrika, 54, pp.567~572 (1967)
- Cox,D.R, "Comment on “Model Uncertainty, Data Mining and Statistical Inference,” by C. Chatfield", Journal of the Royal Statistical Society, 158, pp.419~466 (1995)
- Cox, D. R, "Theoretical Statistics" (1974), Chapman & Hall,London
- Doob,J.L, "Stochastic Processes" (1953), Wiley, NY
- Friedman,M, "Essays in Positive Economics" (1953), University of Chicago Press, Chicago
- Geary,R.C, "The Distribution of Student’s t Ratio for Non-Normal Samples", Journal of the Royal Statistical Society, 3, pp.178~184 (1936)
- Greene,W.H, "Econometric Analysis" (2003), Prentice Hall, NJ
- Hettmansperger,T.P, "Statistical Inference Based on Ranks" (1984), Wiley, NY
- Lehmann,E.L, "Testing Statistical Hypotheses" (1986), Wiley, NY
- Kennedy,P, "Guide to Econometrics" (2008), MIT Press, MA
- Mayo, D. G, "Methodology in Practice: Statistical Misspecification Testing", Philosophy of Science, 71, pp.1007~1025 (2004)
- Mayo, D. G, "Severe Testing as a Basic Concept in a Neyman–Pearson Philosophy of Induction", British Journal for the Philosophy of Science, 57, pp.323~57 (2006)
- Morgan,M.S, "The History of Econometric Ideas" (1990), Cambridge University Press, Cambridge
- Pearson,E.S, "The Analysis of Variance in Cases of Non-Normal Variation", Biometrika, 23, pp.114~133 (1931)
- Pearson,K, "The Fundamental Problem of Practical Statistics", Biometrika, XIII, pp.1~16 (1920)
- Scheffe,H, "The Analysis of Variance" (1959), Wiley, NY.
- Spanos,A, "Statistical Foundations of Econometric Modelling" (1986), Cambridge University Press, Cambridge
- Spanos,A, "On Re-Reading Haavelmo: a Retrospective View of Econometric Modeling", Econometric Theory, 5, pp.405~429 (1989)
- Spanos,A, "The Simultaneous Equations Model Revisited: Statistical Adequacy and Identification", Journal of Econometrics, 44, pp.87~108 (1990)
- Spanos,A, "On Theory Testing in Econometrics: Modeling with Nonexperimental Data", Journal of Econometrics, 67, pp.189~226 (1995)
- Spanos,A, "Probability Theory and Statistical Inference: Econometric Modeling with Observational Data" (1999), Cambridge University Press,Cambridge
- Spanos,A, "Revisiting Data Mining: ‘Hunting’ with or Without a License", The Journal of Economic Methodology, 7, pp.231~264 (2000)
- Spanos,A, "Parametric Versus Non-Parametric Inference: Statistical Models and Simplicity", pp.181~206 (2001), Cambridge University Press
- Spanos,A, "Econometrics in Retrospect and Prospect", New Palgrave Handbook of Econometrics, 1, pp.3~58 (2006)
- Spanos,A, "Curve-Fitting, the Reliability of Inductive Inference and the Error-Statistical Approach", Philosophy of Science, 74, pp.1046~1066 (2007)
- Spanos,A, "Theory Testing in Economics and the Error StatisticalPerspective" (2010), Cambridge University Press, Cambridge
- Spanos,A, "The Model Specification Problem from a Probabilistic Reduction Perspective", Journal of the American Agricultural Association, 83, pp.1168~1176 (2001)
- Staudte, R. G, "Robust Estimation and Testing" (1990), Wiley,NY
- Student, "The Probable Error of the Mean", Biometrika, 6, pp.1~25 (1908)
- Bartlett,M.S, "Some Aspects of the Time-Correlation Problem in Regard to Tests of Significance", Journal of the Royal Statistical Society, 98, pp.536~543 (1935)
- Mayo,D.G, "Error and the Growth of Experimental Knowledge", THE KOREAN ECONOMIC REVIEW, 25(2) (1996)
- Neyman, J, "On the Use and Interpretation of Certain Test Criteria for Purposes of Statistical Inference", Biometrika, 20, pp.175~240 (1928)
- "The Pre-Eminence of Theory vs. the European CVAR ARIS SPANOS: STATISTICAL MISSPECIFICATION AND THE RELIABILITY OF INFERENCE 213 Perspective in Macroeconometric Modeling"[URL]