This text provides a critique of the subjective Bayesian view of statistical inference, and proposes the author's own error-statistical approach as an alternative framework for the epistemology of experiment. It seeks to address the needs of researchers who work with statistical analysis.

H. B. D. Kettlewell's (1955, 1956) classic field experiments on industrial melanism in polluted and unpolluted settings using the peppered moth, Biston betularia, are routinely cited as establishing that the melanic (dark) form of the moth rose in frequency downwind of industrial centers because of the cryptic advantage dark coloration provides against visual predators in soot-darkened environments. This paper critiques three common myths surrounding these investigations: (1) that Kettlewell used a model that identified crypsis as the only selective force responsible (...) for the spread of the melanic gene, (2) that Kettlewell's field experiments alone established that selection for crypsis was the most important factor in the spread of melanic forms, and (3) that Kettlewell's investigations in an unpolluted wood near Dorset constituted a control for his earlier Birmingham studies (contra Hagen 1993, 1996). This analysis further identifies two features that distinguish manipulative experiments in evolutionary biology from experiments in other contexts. First, experiments in evolutionary biology rest on a wealth of information provided by strictly observational ecological studies; in the absence of such information experiments in evolutionary biology make no sense. Second, there is a trade-off between how much control investigators have over the conditions being studied and how informative the results of the experiment will be with regard to natural populations. (shrink)

: Most work done in philosophy of experiment has focused on experiments taken from the domain of physics. The present essay tests whether Allan Franklin's (1984, 1986, 1989, 1990) philosophy of experiment developed in the context of high energy physics can be extended to include examples from evolutionary biology, such as H. B. D. Kettlewell's (1955, 1956, 1958) famous studies of industrial melanism in the peppered moth, Biston betularia. The analysis demonstrates that many of the techniques used by evolutionary biologists (...) exemplify the strategies Franklin lists, and identifies an additional strategy that can likewise be justified by appeal to Bayes's Theorem. (shrink)

A toast is due to one who slays Misguided followers of Bayes, And in their heart strikes fear and terror With probabilities of error! (E.L. Lehmann).

The error statistical account of testing uses statistical considerations, not to provide a measure of probability of hypotheses, but to model patterns of irregularity that are useful for controlling, distinguishing, and learning from errors. The aim of this paper is (1) to explain the main points of contrast between the error statistical and the subjective Bayesian approach and (2) to elucidate the key errors that underlie the central objection raised by Colin Howson at our PSA 96 Symposium.

The Bayesian theory is outlined and its status as a logic defended. In this it is contrasted with the development and extension of Neyman-Pearson methodology by Mayo in her recently published book (1996). It is shown by means of a simple counterexample that the rule of inference advocated by Mayo is actually unsound. An explanation of why error-probablities lead us to believe that they supply a sound rule is offered, followed by a discussion of two apparently powerful objections to the (...) Bayesian theory, one concerning old evidence and the other optional stopping. (shrink)

In this paper, I present a simple and straightforward logic of induction: a consequence relation characterized by a proof theory and a semantics. This system will be called LI. The premises will be restricted to, on the one hand, a set of empirical data and, on the other hand, a set of background generalizations. Among the consequences will be generalizations as well as singular statements, some of which may serve as predictions and explanations.

What role have experiments played, and should they play, in physics? How does one come to believe rationally in experimental results? The Neglect of Experiment attempts to provide answers to both of these questions. Professor Franklin's approach combines the detailed study of four episodes in the history of twentieth century physics with an examination of some of the philosophical issues involved. The episodes are the discovery of parity nonconservation in the 1950s; the nondiscovery of parity nonconservation in the 1930s, when (...) the results of experiments indicated, at least in retrospect, the symmetry violation, but the significance of those results was not realized; the discovery and acceptance of CP symmetry; and Millikan's oil-drop experiment. Franklin examines the various roles that experiment plays, including its role in deciding between competing theories, confirming theories, and calling fo new theories. The author argues that one can provide a philosophical justification for these roles. He contends that if experiment plays such important roles, then one must have good reason to believe in experimental results. He then deals with deveral problems concerning such reslults, including the epistemology of experiment, how one comes to believe rationally in experimental results, the question of the influence of theoretical presuppositions on results, and the problem of scientific fruad. This original and important contribution to the study of the philosophy of experimental science is an outgrowth of many years of research. Franklin brings to this work more than a decade of experience as an experimental high-energy physicist, along with his significant contributions to the history and philosophy of science. (shrink)

This study presents a historical account of experiments, performed by the Collider Detector at Fermilab collaboration, which led to the discovery of the top quark, and a discussion of philosophical issues raised by that episode. ;The historical discussion is based on published and unpublished documents and oral history interviews, and is presented in two parts: First, the formation of the collaboration and construction of the detector are described. The activities of the collaborators during the period of detector construction are described (...) in terms of the development of resources for a general experimental programme. Second, the development of the means of analyzing the data for the top quark search is described, particularly aspects of the analysis that were disputed. The hypothesis that collaboration researchers have come to regard the social process of resolving disputes as a matter of methodological importance is suggested. ;The philosophical discussion of the experiment employs the hierarchy of models approach of Patrick Suppes and Deborah Mayo in order to examine the logic of hypothesis testing and draw some conclusions regarding the nature of scientific evidence. In an extension of an argument presented by Peter Achinstein, the account of hypothesis testing given by hypothetico-deductivist philosophers such as Karl Popper and R. B. Braithwaite is examined in light of the reasoning employed in the top search, and is found wanting. The prediction based on the hypothesis being tested in the top search is found to have been inferred inductively from the experimental data. ;Finally, a discussion is presented of tuning on the signal, a form of bias in the testing of hypotheses. The proscription of this form of bias resembles John Worrall's requirement of use novelty, but is shown instead to serve the aim of devising a test of the hypothesis that is severe, in the sense articulated by Deborah Mayo. It is shown that the evaluation of evidence claims, as it depends on severity considerations, sometimes, as in the case of the top search, requires consideration of the procedures used in designing the test, which may involve historical and social facts about the experimenters and their actions. (shrink)