Abstract
During the 20th century, population ecology and science in general relied on two very different statistical paradigms to solve its inferential problems: error statistics (also referred to as classical statistics and frequentist statistics) and Bayesian statistics. A great deal of good science was done using these tools, but both schools suffer from technical and philosophical difficulties. At the turning of the 21st century (Royall in Statistical evidence: a likelihood paradigm. Chapman & Hall, London, 1997; Lele in The nature of scientific evidence: statistical, philosophical and empirical considerations. The University of Chicago Press, Chicago, pp 191–216, 2004a), evidential statistics emerged as a seriously contending paradigm. Drawing on and refining elements from error statistics, likelihoodism, Bayesian statistics, information criteria, and robust methods, evidential statistics is a statistical modern synthesis that smoothly incorporates model identification, model uncertainty, model comparison, parameter estimation, parameter uncertainty, pre-data control of error, and post-data strength of evidence into a single coherent framework. We argue that evidential statistics is currently the most effective statistical paradigm to support 21st century science. Despite the power of the evidential paradigm, we think that there is no substitute for learning how to clarify scientific arguments with statistical arguments. In this paper we sketch and relate the conceptual bases of error statistics, Bayesian statistics and evidential statistics. We also discuss a number of misconceptions about the paradigms that have hindered practitioners, as well as some real problems with the error and Bayesian statistical paradigms solved by evidential statistics.
Similar content being viewed by others
Notes
A “model form” is synonymous with what Taper (2004) and Taper et al. (2008) have called a “model structure”. We are changing our vocabulary to avoid collision with the literature use of “model structure” as a broader term implying on a topological relationship among model elements without specifying either functional form of relationships or parameter values.
Statistics uses two different symbols (; and |) to indicate conditioning: \(\Pr \left( {x;A} \right)\) indicates that the probability of x is being conditioned on a fixed value, while \(\Pr \left( {x|B} \right)\) indicates that the probability of x is being conditioned on a random variable. In casual use, these symbols are sometimes used interchangeably, or omitted.
When the abbreviation IC refers to an information criterion as a procedure or algorithm it will be given in Roman typeface. When IC refers to a value calculated from data, it will be set in italic.
Note Lele’s parameter a is equivalent to our r i used above.
References
Aho K, Derryberry D, Peterson T (2014) Model selection for ecologists: the worldviews of AIC and BIC. Ecology 95:631–636
Akaike H (1973) Information theory as an extension of the maximum likelihood principle. In: Petrov BN, Csaki F (eds) second international symposium on information theory. Akademiai Kiado, Budapest, pp 267–281
Akaike H (1974) A new look at statistical-model identification. IEEE Trans Autom Control 19:716–723
Bandyopadhyay PS, Brittan G, Taper ML (2016) Belief, evidence, and uncertainty: problems of epistemic inference. SpringerBriefs in Philosophy of Science. Springer, Cham (in press)
Barnard GA (1949) Statistical inference. J R Stat Soc Ser B 11:115–149
Barnard GA (1967) The use of the likelihood function in statistical practice. In: Le Cam L, Neyman J (eds) Proceedings of the fifth berkeley symposium. University of California Press, Berkeley, pp 27–40
Barnett V (1999) Comparative statistical inference, 3rd edn. Wiley, Chinchester
Basu A, Shioya H, Park C (2011) Statistical inference: the minimum distance approach. CRC Press, Boca Raton
Beaumont MA, Rannala B (2004) The Bayesian revolution in genetics. Nat Rev Genet 5:251–261
Blume J, Peipert JF (2003) What your statistician never told you about P-values. J Am Assoc Gynecol Laparosc 10:439–444
Box GEP (1976) Science and statistics. J Am Stat Assoc 71:791–799
Bozdogan H (1987) Model selection and Akaike information criterion (AIC): the general theory and its analytical extensions. Psychometrika 52:345–370
Burnham KP, Anderson DR (2002) Model selection and multi-model inference: a practical information-theoretic approach, 2nd edn. Springer, New York
Burnham KP, Anderson DR (2004) Multimodel inference—understanding AIC and BIC in model selection. Soc Method Res 33:261–304
Chatfield C (1995) Model uncertainty, data mining and statistical inference. J R Stat Soc Ser A 158:419–466
Christen JA, Nakamura M (2000) On the analysis of accumulation curves. Biometrics 56:748–754
Clark JS (2005) Why environmental scientists are becoming Bayesians. Ecol Lett 8:2–14
Cohen JE (2004) Mathematics is biology’s next microscope, only better; biology is mathematics’ next physics, only better. PLoS Biol 2:2017–2023
Dennis B, Ponciano JM (2014) Density-dependent state-space model for population-abundance data with unequal time intervals. Ecology 95:2069–2076
Dennis B, Ponciano JM, Lele SR, Taper ML, Staples DF (2006) Estimating density dependence, process noise, and observation error. Ecol Monogr 76:323–341
Dorazio RM (2015) Bayesian data analysis in population ecology: motivations, methods, and benefits. Popul Ecol. doi:10.1007/s10144-015-0503-4
Edwards AWF (1992) Likelihood, expanded. Johns Hopkins University Press, Baltimore
Efron B (2010) Large-scale inference: empirical Bayes methods for estimation, testing, and prediction Institute of mathematical statistics monographs. Cambridge Univ. Press, Cambridge
Efron B (2013) Bayes’ theorem in the 21st century. Science 340:1177–1178
Ellison AM (2004) Bayesian inference in ecology. Ecol Lett 7:509–520
Fisher RA (1922) On the mathematical foundations of theoretical statistics. Philos Trans R Soc Lond Ser A 222:309–368
Fisher RA (1956) Statistical methods and scientific inference. Oliver and Boyd, London
Fisher RA (1971) The design of experiments, 8th edn. Hafner Publishing Company, New York
Frigg R (2006) Scientific representation and the semantic view of theories. Theoria 55:49–65
Gause GF (1934) The struggle for existence. Williams and Wilkins, Baltimore
Gelman A, Hennig C (2015) Beyond subjective and objective in statistics. Columbia University Department of Statistics technical report. http://www.stat.columbia.edu/~gelman/research/unpublished/objectivity13.pdf. Accessed 24 Feb 2015
Gelman A, Shalizi CR (2013) Philosophy and the practice of Bayesian statistics. Br J Math Stat Psychol 66:8–38
Giere R (1988) Explaining science. University of Chicago Press, Chicago
Giere RN (1999) Science without laws (science and its conceptual foundations). University of Chicago Press, Chicago
Giere RN (2004) How models are used to represent reality. Philos Sci 71:742–752
Giere RN (2008) Models, metaphysics, and methodology. In: Hartmann S, Bovens L, Hoefer C (eds) Nancy Cartwright’s philosophy of science. Routledge, New York
Gimenez O, Buckland ST, Morgan BJT, Bez N, Bertrand S, Choquet R, Dray S, Etienne M-P, Fewster R, Gosselin F, Merigot B, Monestiez P, Morales JM, Mortier F, Munoz F, Ovaskainen O, Pavoine S, Pradel R, Schurr FM, Thomas L, Thuiller W, Trenkel V, de Valpine P, Rexstad E (2014) Statistical ecology comes of age. Biol Lett 10:20140698
Guttorp P (1995) Stochastic modeling of scientific data. Chapman & Hall, London
Hacking I (1965) Logic of statistical inference. Cambridge University Press, Cambridge
Hájek A (2012) Interpretations of probability. In: Zalta EN (ed) The Stanford encyclopedia of philosophy (winter 2012 edition). http://plato.Stanford.Edu/archives/win2012/entries/probability-interpret/. Stanford
Hannan EJ, Quinn BG (1979) Determination of the order of an autoregression. J R Stat Soc Ser B 41:190–195
Hughes RIG (1997) Models and representation. Philos Sci Proc 64:325–336
Hurvich CM, Tsai CL (1989) Regression and time series model selection in small samples. Biometrika 76:297–307
Kalbfleisch JG (1985) Probability and statistical inference. Volume ii: statistical inference, 2nd edn. Springer, New York
Kass RE, Raftery AE (1995) Bayes factors. J Am Stat Assoc 90:773–795
Kolmogorov AN (1933) Grundbegriffe der wahrscheinlichkeitrechnung, ergebnisse der mathematik; translated as Foundations of probability (1950). Chelsea Publishing Company, New York
Konishi S, Kitagawa G (2008) Information criteria and statistical modeling. Springer, New York
Kuhnert PM, Martin TG, Griffiths SP (2010) A guide to eliciting and using expert knowledge in Bayesian ecological models. Ecol Lett 13:900–914
Lele SR (2004a) Evidence functions and the optimality of the law of likelihood. In: Taper ML, Lele SR (eds) The nature of scientific evidence: statistical, philosophical and empirical considerations. The University of Chicago Press, Chicago, pp 191–216
Lele SR (2004b) Elicit data, not prior: on using expert opinion in ecological studies. In: Taper ML, Lele SR (eds) The nature of scientific evidence: statistical, philosophical and empirical considerations. The University of Chicago Press, Chicago, pp 410–435
Lele SR (2015) Is non-informative Bayesian analysis appropriate for wildlife management: survival of San Joaquin kit fox and declines in amphibian populations. arXiv preprint arXiv:150200483
Lele SR, Allen KL (2006) On using expert opinion in ecological analyses: a frequentist approach. Environmetrics 17:683–704
Lele SR, Taper ML (2012) Information criteria in ecology. In: Hastings A, Gross L (eds) Encyclopedia of theoretical ecology. University of California Press, Berkley, pp 371–376
Lele SR, Dennis B, Lutscher F (2007) Data cloning: easy maximum likelihood estimation for complex ecological models using Bayesian Markov chain Monte Carlo methods. Ecol Lett 10:551–563
Lele SR, Nadeem K, Schmuland B (2010) Estimability and likelihood inference for generalized linear mixed models using data cloning. J Am Stat Assoc 105:1617–1625
Lindblom CE (1959) The science of muddling through. Publ Admin Rev 19:79–88
Lindley DV (1957) A statistical paradox. Biometrika 44:187–192
Lindley DV (2000) The philosophy of statistics. J R Stat Soc Ser D 49:293–337
Lindsay BG (2004) Statistical distances as loss functions in assessing model adequacy. In: Taper ML, Lele SR (eds) The nature of scientific evidence: statistical, philosophical and empirical considerations. The University of Chicago Press, Chicago, pp 439–488
Mayo DG (1996) Error and the growth of experimental knowledge. University of Chicago Press, Chicago
Mayo DG, Cox DR (2006) Frequentist statistics as a theory of inductive inference. In: Optimality: The 2nd Lehmann symposium. Institute of Mathematical Statistics, Lecture Notes—Monograph Series, vol 57, pp 77–97
Mayo DG, Spanos A (2006) Severe testing as a basic concept in a Neyman–Pearson philosophy of induction. Br J Philos Sci 57:323–357
Montoya JA (2008) La verosimilitud perfil en la inferencia estadística. Doctoral Dissertation, Center for Research in Mathematics, Guanajuato, México (in Spanish)
Montoya JA, Diaz-Frances E, Sprott DA (2009) On a criticism of the profile likelihood function. Stat Papers 50:195–202
Moreno M, Lele SR (2010) Improved estimation of site occupancy using penalized likelihood. Ecology 91:341–346
Morgan M (1999) Learning from models. In: Morrison M, Morgan M (eds) Models as mediators: perspectives on natural and social science. Cambridge University Press, Cambridge, pp 347–388
Newman KB, Buckland ST, Morgan BJT, King RS, Borchers DL, Cole DJ, Thomas L (2014) Modeling population dynamics. Springer, New York
Neyman J, Pearson ES (1933) On the problem of the most efficient tests of statistical hypostheses. Philos Trans R Soc Lond Ser A 231:289–337
Pawitan Y (2001) In all likelihood: statistical modeling and inference using likelihood. Oxford University Press, Oxford
Peirce CS (1878) Illustrations of the logic of science iii—the doctrine of chances. Popul Sci Mon 12:604–615
Ponciano JM, Taper ML, Dennis B, Lele SR (2009) Hierarchical models in ecology: confidence intervals, hypothesis testing, and model selection using data cloning. Ecology 90:356–362
Ponciano JM, Burleigh G, Braun EL, Taper ML (2012) Assessing parameter identifiability in phylogenetic models using data cloning. Syst Biol 61:955–972
Popper KR (1959) The propensity interpretation of probability. Br J Philos Sci 10:25–42
Raftery AE (1995) Bayesian model selection in social research. Sociol Methodol 25:111–163
Rannala B (2002) Identifiability of parameters in MCMC Bayesian inference of phylogeny. Syst Biol 51:754–760
Rice JA (1995) Mathematical statistics and data analysis, 2nd edn. Duxbury Press, Belmont
Royall RM (1986) The effect of sample-size on the meaning of significance tests. Am Stat 40:313–315
Royall RM (1997) Statistical evidence: a likelihood paradigm. Chapman & Hall, London
Royall RM (2000) On the probability of observing misleading statistical evidence. J Am Stat Assoc 95:760–780
Royall RM (2004) The likelihood paradigm for statistical evidence. In: Taper ML, Lele SR (eds) The nature of scientific evidence: statistical, philosophical and empirical considerations. The University of Chicago Press, Chicago, pp 119–152
Royle JA, Dorazio RM (2008) Hierarchical modeling and inference in ecology: the analysis of data from populations, metapopulations and communities. Academic Press, San Deigo
Schwarz G (1978) Estimating the dimension of a model. Ann Stat 6:461–464
Solymos P (2010) dClone: data cloning in R. R J 2:29–37
Sprott DA (2000) Statistical inference in science. Springer, New York
Sprott DA (2004) What is optimality in scientific inference? In: Rojo J, PerezAbreu V (eds) First Erich L. Lehmann symposium—optimality. Institute of Mathematical Statistics, Lecture Notes—Monograph Series, vol 44, pp 133–152
Strevens M (2010) Reconsidering authority: scientific expertise, bounded rationality, and epistemic backtracking, Chap. 13. In: Gendler TS, Hawthorne J (eds) Oxford studies in epistemology, vol 3. Oxford University Press, Oxford, pp 294–330
Suppe F (1989) The semantic conception of theories and scientific realism. University of Chicago Press, Chicago
Taper ML (2004) Model identification from many candidates. In: Taper ML, Lele SR (eds) The nature of scientific evidence: statistical, philosophical and empirical considerations. The University of Chicago Press, Chicago, pp 448–524
Taper ML, Lele SR (eds) (2004) The nature of scientific evidence: statistical, philosophical and empirical considerations. The University of Chicago Press, Chicago
Taper ML, Lele SR (2011) Evidence, evidence functions, and error probabilities. In: Bandyopadhyay PS, Forster MR (eds) Philosophy of statistics. Elsevier, Oxford, pp 513–532
Taper ML, Ponciano JM (2016) Projections in model space: multimodel inference beyond model averaging. In: Bandyopadhyay P, Brittan G, Taper ML (eds) Belief, evidence, and uncertainty: problems of epistemic inference. SpringerBriefs in Philosophy of Science. Springer, Cham (in press)
Taper ML, Staples DF, Shepard BB (2008) Model structure adequacy analysis: selecting models on the basis of their ability to answer scientific questions. Synthese 163:357–370
Thompson B (2007) The nature of statistical evidence. Springer, New York
Underwood AJ (1997) Experiments in ecology: their logical design and interpretation using analysis of variance. Cambridge University Press, Cambridge
van der Tweel I (2005) Repeated looks at accumulating data: To correct or not to correct? Eur J Epidemiol 20:205–211
van Fraassen B (2002) The empirical stance. Yale University Press, New Haven
Venn J (1876) The logic of chance, 2nd edn reprinted 1962. Chelsea Publishing Co., New York
von Mises R (1951) Probability, statistics, and truth, 3rd edn (English translation 1957). George Allen & Unwin Ltd, London
Walker AM (1969) On asymptotic behaviour of posterior distributions. J Roy Stat Soc Ser B 31:80–88
Wang JPZ, Lindsay BG (2005) A penalized nonparametric maximum likelihood approach to species richness estimation. J Am Stat Assoc 100:942–959
Wilks SS (1938) The large-sample distribution of the likelihood ratio for testing composite hypotheses. Ann Math Stat 9:60–62
Yamamura K (2015) Bayes estimates as an approximation to maximum likelihood estimates. Popul Ecol. doi:10.1007/s10144-015-0526-x
Acknowledgments
We thank Dr. Yukihiko Toquenaga for inviting MLT to present in a plenary symposium of the 30th Annual Meeting of the Society of Population Ecology in Tsukuba, Japan. We are also grateful to the Society and to the Japan Society for the Promotion of Science for providing funding. MLT was partially supported by US National Science Foundation grant # DUE-1432577. JMP was partially supported by US National Institute of Health grant # R01 GM103604. We thank Ian Ausprey, Juan Pablo Gomez, Brian Dennis, and Robert Holt for insightful comments and useful suggestion helping to improve this manuscript. We also would like to thank Jack Sullivan for his questions about information criteria, and Tessa Barton for her questions about the subjectivity of model choice. MLT would like to thank Prasanta Bandyopadhyay and Gordon Brittan for many discussions on the philosophy of statistics during the production of Bandyopadhyay et al. 2016. This paper and that work were produced simultaneously and ideas have filtered between the two. The authors wish to thank also the constructive critiques of Michael J. Lew and another anonymous reviewer. We thank the University of California Press for permission to reprint Subhash R. Lele, Mark L. Taper “Information Criteria in Ecology”, 371–375, Figure 1, by Hastings, Alan and Louis J. Gross in Encyclopedia of Theoretical Ecology (c) 2012 by the Regents of the University of California. Published by the University of California Press. We thank the Journal of the American Statistical Association for permission to reprint Richard M. Royall (2000). On the probability of observing misleading statistical evidence. J Amer Stat Assoc 95:760–780, Figure 2. We are grateful to Mayuko Tanigawa and Yukihiko Toquenaga for editorial patience in helping us get things right despite extreme platform translation problems.
Author information
Authors and Affiliations
Corresponding author
Additional information
This manuscript was submitted for the special feature based on a symposium in Tsukuba, Japan, held on 11 October 2014.
Rights and permissions
About this article
Cite this article
Taper, M.L., Ponciano, J.M. Evidential statistics as a statistical modern synthesis to support 21st century science. Popul Ecol 58, 9–29 (2016). https://doi.org/10.1007/s10144-015-0533-y
Received:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s10144-015-0533-y