Abstract
We will start this chapter with a brief exposition of HD(hypothetico-deductive)testing, that is, the HD-method of testing hypotheses, and indicate the related qualitative explication of confirmation as well as quantitative extensions. HD-testing attempts to give an answer to one of the questions that one may be interested in, the truth question, which may be qualified according to the relevant epistemological position.1 However, the (theory) realist, for instance, is not only interested in the truth question, but also in some other questions. To begin with, there is the more refined question of which (individual or general) facts2 the hypothesis explains (its explanatory successes) and with which facts it is in conflict (its failures); the success question for short. We will show in this chapter that the HD-method can also be used in such a way that it is functional for (partially) answering this question. This method is called HD-evaluation, and uses HD-testing. Since the realist ultimately aims to approach the strongest true hypothesis, if any, i.e. the (theoretical-cum-observational) truth about the subject matter, the plausible third aim of the HD-method is to help answer the question of how far a hypothesis is from the truth, the truth approximation question Here the truth will be taken in a relatively modest sense, viz. relative to a given domain and conceptual frame. In the next chapter we will make plausible that HD-evaluation is also functional for answering the truth approximation question.
This is a preview of subscription content, log in via an institution to check access.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
Preview
Unable to display preview. Download preview PDF.
Reference
Recall from Subsection 2.4.3 that we use in this book the term ‘epistemological position’ in a broad sense, as a view concerning the nature of (claims to) knowledge, including relevant ontological and semantic presuppositions. ’Methodology’ is conceived as the way in which claims to knowledge are evaluated in terms of evidence.
When we speak of individual and general facts we assume that these facts are formulated in observation terms and that the hypotheses describing them have been sufficiently confirmed to be accepted. Of course, we may nevertheless be mistaken, we just assume for the time being that we are not.
If one finds the term ‘counter-example’ to have a realist, or falsificationist, flavor, one may replace it systematically by ’problem’ or ’failure’.
The distinction between testing and evaluation somewhat corresponds to Schaffner’s distinction between local and global evaluation (Schaffner, 1993, Chapter V), but the precise relation is difficult to indicate.
McAllister (1996) convincingly argues for three main claims about plausibility considerations, at least as far as they are of an, in a broad sense, aesthetic nature. First, in normal science they govern the preferences between equally successful theories, and only between such theories. Second, they originate from ‘aesthetic induction’ applied to the accepted background knowledge. Three, the opposing parties in scientific revolutions arise from the strong differences in the willingness to sacrifice dominant aesthetic considerations to empirical success.
This follows directly from the general definition of conditional probability, viz. p(A/B) = aef p{AundB)/p(B), according to which p{E/H)/p{E) =p{EundH)/(p{H)p(E)) =p(H/E)/ p(H)
The P2-related arguments concern the first and the second argument in Fitelson’s Table 1, and the second in Table 2. Of the other two, the example of ‘unintuitive’ confirmation is rebutted in ICR (Chapter 3) with a similar case against the difference measure. The other one is related to the ’grue-paradox’, for which Chapter 2 and 3 of ICR claim to present an illuminating analysis in agreement with P2.
Harre ( 1993, Chapter 3) distinguishes three dimensions of generality: substance, experimental, spatio-temporal, and their possible restrictions. We take Harre’s first and third dimension into account in the domain, and the second dimension in the initial conditions (see below).
Here we neglect the famous asymmetry problems about causal explanations, in particular concerning individual facts. The standard example is that the length of a flag pole cannot be explained by the length of its shadow. Note, however, that this problem does not arise when general facts are involved. For example, the proportionality between the length of both can be explained by the classic theory of light propagation.
Note that the above treatments of ‘C0 and not-F 0 ’ and ‘C0 and F0’ and the ones to come about ’not-C 0 and not-Fo’ and ’not-C 0 and F0’ are essentially in agreement with the way of dealing with (non-)black (non-)ravens, in the context of testing the raven hypothesis in Chapter 7.
For a survey of classical and Bayesian approaches to statistical testing, see Schaffner (1993, Chapter V). For a more detailed presentation, see Howson and Urbach (1989).
Informally: for all sufficiently large random samples 5 of individuals from domain D satisfying condition C, it holds, with high probability, that the ratio of individuals satisfying F is in a certain region.
Informally: for all sufficiently large samples si and s2 of individuals of domains D1 and D2,satisfying C, it holds, with high probability, that the ratio in si satisfying F is significantly larger than the percentage in s2 satisfying F.
Author information
Authors and Affiliations
Rights and permissions
Copyright information
© 2001 Springer Science+Business Media Dordrecht
About this chapter
Cite this chapter
Kuipers, T.A.F. (2001). Testing and Further Separate Evaluation of Theories. In: Structures in Science. Synthese Library, vol 301. Springer, Dordrecht. https://doi.org/10.1007/978-94-015-9739-5_7
Download citation
DOI: https://doi.org/10.1007/978-94-015-9739-5_7
Publisher Name: Springer, Dordrecht
Print ISBN: 978-90-481-5749-5
Online ISBN: 978-94-015-9739-5
eBook Packages: Springer Book Archive