Abstract
The econometrics literature is dominated by the frequentist school which places primary emphasis on the specification of methods that have certain long-run frequency properties, mostly disavowing any notion of inference based on the given data. This preference for frequentism is at least partially based on the belief that probabilistic inference is possible only through a Bayesian approach, the success of which generally depends on the unrealistic assumption that the prior distribution is meaningful in some way. This paper is intended to inform econometricians that an alternative inferential model (IM) approach exists that can achieve probabilistic inference without a prior and while enjoying certain calibration properties essential for reproducibility, etc. Details about the IM construction and its properties are presented, along with some intuition and examples.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
Notes
- 1.
At his keynote address for a workshop on foundations at Rutgers University in April 2016, http://statistics.rutgers.edu/bff2016.
- 2.
This is not an ideal choice of word because “predict” has a particular meaning in statistics and econometrics, but the meaning here is different. “Guess” or “impute” are other potential words to describe the operation in consideration, but both still miss the mark slightly. A more accurate description of what I have in mind is “to cast a net” at the target.
- 3.
The formula (5) silently assumes that \(\varTheta _y(\mathcal {S})\) is non-empty with \(\mathsf {P}_\mathcal {S}\)-probability 1 for all y. This holds automatically in many cases, often because of the preprocessing discussed in Sect. 3, but not in all. The ideal remedy seems to be choosing a predictive random set which is elastic in a certain sense; see Ermini Leaf and Liu [12].
- 4.
- 5.
- 6.
Efficiency is essentially the rival to validity, and the goal is to balance between the two. For example, a belief function that always takes value 0 for all \(A \ne \varTheta \), which is achieved by a choosing an extreme predictive random set \(\mathcal {S}\equiv \mathbb {U}\), is sure to be valid, but then the corresponding plausibility function would always be 1; consequently, the plausibility regions would be unbounded and practically useless. So, roughly, the idea behind efficiency is to have the smallest predictive random set such that the IM is still efficient. More on this in Sect. 5.
References
Berger J (2006) The case for objective Bayesian analysis. Bayesian Anal 1(3):385–402
Cheng Q, Gao X, Martin R (2014) Exact prior-free probabilistic inference on the heritability coefficient in a linear mixed model. Electron J Stat 8(2):3062–3076
Creasy MA (1954) Symposium on interval estimation: limits for the ratio of means. J R Stat Soc Ser B 16:186–194
Dempster AP (1966) New methods for reasoning towards posterior distributions based on sample data. Ann Math Stat 37:355–374
Dempster AP (1967) Upper and lower probabilities induced by a multivalued mapping. Ann Math Stat 38:325–339
Dempster AP (1968) A generalization of Bayesian inference (With discussion). J R Stat Soc Ser B 30:205–247
Dempster AP (2008) The Dempster-Shafer calculus for statisticians. Int J Approx Reason 48(2):365–377
Dempster AP (2014) Statistical inference from a Dempster–Shafer perspective. In: Lin X, Genest C, Banks DL, Molenberghs G, Scott DW, Wang J-L (eds) Past, present, and future of statistical science. Chapman & Hall/CRC Press. Chap 24
Dufour J-M (1997) Some impossibility theorems in econometrics with applications to structural and dynamic models. Econometrica 65(6):1365–1387
Efron B (1998) R. A. Fisher in the 21st century. Stat Sci 13(2):95–122
Efron B (2013) Discussion: “Confidence distribution, the frequentist distribution estimator of a parameter: a review” [mr3047496]. Int Stat Rev 81(1):41–42
Ermini Leaf D, Liu C (2012) Inference about constrained parameters using the elastic belief method. Int J Approx Reason 53(5):709–727
Fidler F, Thomason N, Cummings G, Fineh S, Leeman J (2004) Editors can lead researchers to confidence intervals, but can’t make them think. Psychol Sci 15:119–126
Fieller EC (1954) Symposium on interval estimation: some problems in interval estimation. J R Stat Soc Ser B 16:175–185
Fisher RA (1973) Statistical methods and scientific inference, 3rd edn. Hafner Press, New York
Fraser DAS (2004) Ancillaries and conditional inference. Stat Sci 19(2):333–369 With comments and a rejoinder by the author
Ghosh JK, Delampady M, Samanta T (2006) An introduction to Bayesian analysis. Springer, New York
Ghosh M (2011) Objective priors: an introduction for frequentists. Stat Sci 26(2):187–202
Gleser LJ, Hwang JT (1987) The nonexistence of \(100(1-\alpha )\)% confidence sets of finite expected diameter in errors-in-variables and related models. Ann Stat 15(4):1351–1362
Hannig J (2009) On generalized fiducial inference. Stat Sinica 19(2):491–544
Hannig J, Iyer H, Lai RCS, Lee TCM (2016) Generalized fiducial inference: a review and new results. J Am Stat Assoc 111(515):1346–1361
Martin R (2015) Plausibility functions and exact frequentist inference. J Am Stat Assoc 110:1552–1561
Martin R (2017) On an inferential model construction using generalized associations. J Stat Plann Infer. doi:10.1016/j.jspi.2016.11.006 (to appear)
Martin R (2017) A statistical inference course based on p-values. Am Stat. doi:10.1080/00031305.2016.1208629 (to appear)
Martin R, Lin Y (2016) Exact prior-free probabilistic inference in a class of non-regular models. Statistics 5(1):312–321
Martin R, Lingham R (2016) Prior-free probabilistic prediction of future observations. Technometrics 58(2):225–235
Martin R, Liu C (2013) Inferential models: a framework for prior-free posterior probabilistic inference. J Am Stat Assoc 108(501):301–313
Martin R, Liu C (2014) A note on \(p\)-values interpreted as plausibilities. Stat Sinica 24:1703–1716
Martin R, Liu C (2015) Conditional inferential models: combining information for prior-free probabilistic inference. J R Stat Soc Ser B 77(1):195–217
Martin R, Liu C (2015) Inferential models: reasoning with uncertainty. Monographs in statistics and applied probability series. Chapman & Hall/CRC Press, Boca Raton
Martin R, Liu C (2015) Marginal inferential models: prior-free probabilistic inference on interest parameters. J Am Stat Assoc 110:1621–1631
Martin R, Liu C (2016) Validity and the foundations of statistical inference. arXiv:1607.05051 (Unpublished manuscript)
Martin R, Xu H, Zhang Z, Liu C (2016) Valid uncertainty quantification about the model in a linear regression setting. arXiv:1412.5139 (Unpublished manuscript)
Molchanov I (2005) Theory of random sets. Probability and its applications (New York). Springer, London
Nguyen HT (1978) On random sets and belief functions. J Math Anal Appl 65(3):531–542
Nguyen HT (2006) An introduction to random sets. Chapman & Hall/CRC, Boca Raton
Nuzzo R (2014) Scientific method: statistical errors. Nature 506:150–152
Schweder T, Hjort NL (2002) Confidence and likelihood. Scand J Stat 29(2):309–332
Schweder T, Hjort NL (2016) Confidence, likelihood, probability: statistical inference with confidence distributions. Cambridge University Press, Cambridge
Shafer G (1976) A mathematical theory of evidence. Princeton University Press, Princeton
Shafer G (1979) Allocations of probability. Ann Probab 7(5):827–839
Shafer G (1982) Belief functions and parametric models. J Roy Stat Soc Ser B 44(3):322–352 With discussion
Shafer G (1987) Belief functions and possibility measures. In: Bezdek JC (ed) The analysis of fuzzy information. Mathematics and logic, vol 1. CRC, pp 51–84
Trafimowa D, Marks M (2015) Editorial. Basic Appl Soc Psychol 37(1):1–2
Xie M, Singh K (2013) Confidence distribution, the frequentist distribution of a parameter—a review. Int Stat Rev 81(1):3–39
Yang R, Berger JO (1998) A catalog of non-informative priors. http://www.stats.org.uk/priors/noninformative/YangBerger1998.pdf (Unpublished manuscript)
Zabell SL (1992) R. A. Fisher and the fiducial argument. Stat Sci 7(3):369–387
Acknowledgements
The author thanks the organizers of the 10th International Conference of the Thailand Econometric Society, in particular, Professor Hung Nguyen, for the invitation to present at the conference and to submit a paper for the proceedings.
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2017 Springer International Publishing AG
About this chapter
Cite this chapter
Martin, R. (2017). Prior-Free Probabilistic Inference for Econometricians. In: Kreinovich, V., Sriboonchitta, S., Huynh, VN. (eds) Robustness in Econometrics. Studies in Computational Intelligence, vol 692. Springer, Cham. https://doi.org/10.1007/978-3-319-50742-2_10
Download citation
DOI: https://doi.org/10.1007/978-3-319-50742-2_10
Published:
Publisher Name: Springer, Cham
Print ISBN: 978-3-319-50741-5
Online ISBN: 978-3-319-50742-2
eBook Packages: EngineeringEngineering (R0)