(1) See Simmons, R.G., Klein, S.D. and Thornton, K., The Family Member's Decision to be a Kidney Transplant Donor, Journal of Comparative Family Studies, IV (1973), 88–115Google Scholar.

(2) Whether an ‘irrational’ assessment is the same as an ‘unscientific’ one is another issue with which I am not concerned in this paper. But I would suggest that the fact that the results of an assessment do not square with the present state of scientific knowledge tells us nothing about the rationality of the procedures used to make that assessment.

(3) Davidson, D., Suppes, P. and Siegel, S., Decision-Making: an experimental approach (Stanford, Stanford University Press, 1957)Google Scholar.

(4) The test is a weak one since it is quite possible that a billiards player would be a competent mathematician (and hence able to solve the problems we set him) but would still fail to use mathematics in selecting his shots. This example comes from Cyert, R. M. and March, J.G., A Behavioral Theory of the Firm (Englewood Cliffs, N. J., Prentice-Hall, 1963), p. 306Google Scholar.

(5) Bass, F.M., Pessemier, E.A. and Lehmann, D.R., An Experimental Study of Relationships Between Attitudes, Brand Preference, and Choice, Behavioral Science, XVII (1972), 532–541CrossRefGoogle Scholar.

(6) In a study of transitivity MacCrimmon confronted subjects with their intransitive choices. Only two of the thirty-eight subjects wished to stand by their original choices: the others all claimed that they had made mistakes. See MacCrimmon, K.R, Descriptive and Normative Implications of the Decision-theory Post-ulates, in Borch, K. and Mossin, J. (eds.), Risk and Uncertainty (London, Macmillan, 1968), pp. 3–23Google Scholar.

(7) Katona, G. and Mueller, E., A Study of Purchase Decisions, in Clark, L. (ed.), Consumer Behavior (New York, New York University Press, 1954), pp. 30–87Google Scholar.

(8) The belief that one needs to take a long time if one is to reach a rational decision is a common error, as is the belief that the man who collects more information is thereby making a more rational decision. There may be nothing irrational whatsoeverin dealing with a big decision very quickly and taking a long time to cope with a small one: the time taken may simply depend on how balanced the various alternatives are.

(9) Edwards, W., Behavioral Decision Theory, Annual Review of Psychology, XIII (1961), p. 477Google Scholar.

(10) Marx, K. and Engels, F., Manifesto of the Communist Party (Moscow, Progress Publishers, 1965), p. 94Google Scholar.

(11) Latané, B. and Darley, J. M., Social Determinants of Bystander Intervention in Emergencies, in Macaulay, J. and Berkowitz, L. (eds.), Altruism and Helping Behavior (New York, Academic Press, 1970), pp. 13–27Google Scholar.

(12) Davenport, W., Jamaican Fishing: a game theory analysis, Yale University Publications in Anthropology, LIX (1960), 3–11Google Scholar For a more detailed critique sec Read, D. W. and Read, C. E., A Critique of Davenport's Game Theory Analysis, American Anthropologist, LXXII (1970), 351–355CrossRefGoogle Scholar.

(13) Davenport also makes various predictions (with equal accuracy) about the proportions of fishermen who will follow various strategies. The matrix for this is more complicated and for ease of exposition I have omitted it.

(14) Setting all one's pots inside the bank is in fact the minimax strategy (if we equatemoney values with utility), although it is not a very sensible strategy to follow against an opponent using a mixed strategy or otherwise varying his behaviour in some fixed ratio. But in any event I see absolutely no reason why we should equate money values with utility.

(15) See Lieberman, B., Human Behavior Behavin a Strictly Determined 3 × 3 Matrix, Behavioral Science, V (1960), 317–322Google Scholar, and Brayer, A. R., An Experimental Analysis of Some Variables of Minimax Theory, Behavin ioral Science, IX (1964), 33–44CrossRefGoogle Scholar.

(16) See MacCrimmon, op. cit.

(17) Popper, K., The Open Society and Its Enemies (London, Routledge and Kegan Paul, 1966), Vol. II, p. 155Google Scholar.

(18) Different subjects had different preference orderings and not everyone put Oxford first. But for the sake of simplicity I have inserted Oxford, Warwick and Leeds into the answers reported in place of the actual Universities mentioned.

(19) I originally gave subjects various lists of universities and asked them to rank them and to state which (in utility terms) came at the midpoint of each list. I then made the ‘certainty’ in the choices given to the subjects a university which came just above or just below this midpoint. This technique yielded 27 correct predictions with the expected utility model out of 39 possible ones. I subsequently presented subjects with various pairs of universities and asked them to rate the difference (on a five point scale) between the two members of each pair. With this technique the expected utility model gave 22 correct predictions out of 25 possible ones.

(20) Barth, F., Models of Social Organi-sation (Royal Anthropological Institute Occasional Paper No. 23, 1966)Google Scholar.

(21) See, for example, Cancian, F., Stratification and Risk-taking: a theory tested on agricultural innovations, American Sociological Review, XXXII (1967), 912–927CrossRefGoogle Scholar.

(22) The economist might however argue that the richer man will be further along his marginal utility curve and at a point where the curve is less steep. This would mean that the gamble has a smaller variance. It could then be argued that the smaller the variance of a gamble the more willin people are to accept it, but I know of no evidence that this is correct.

(23) See Becker, W. and Siegel, S., Utility of Grades: level of aspiration in a decision theory context, Journal of Expe-rimental Psychology, LV (1958), 81–85CrossRefGoogle Scholar.

(24) In my explanation I have assumed that the richer farmers are the ones who have been more successful in the past and/or have higher aspiration levels, and that farmers are concerned with their relative, not absolute, success. In the case of the skip pers, Barth reports, both these assumptions are correct.

(25) By far the most convincing demon-stration of an interval scale of utility is that by Davidson, Suppes and Siegel (op. cit.). Their technique, however, requires a continuous variable, which places consider-able restrictions on its applicability. Fur-thermore, they carry out only very weak tests to validate that it is in fact an interval scale which they have obtained, and four of their nineteen subjects fail even these weak tests.

(26) The fact that one assigns numbers like ‘99 % certain’ does not necessarily mean that a ratio scale has been obtained. To show that we do in fact have a ratio scale it is essential to demonstrate that there are empirical operations that can be performed which are isomorphic with the mathematical operations that can be carried out with a ratio scale.

(27) See Simon, H.A., Models of Man (New York, Wiley, 1957)Google Scholar.

(28) The view of decision-making that I am putting forward here is essentially a dynamic one: the individual first considers the main options open to him and, only if he cannot reach a decision here, goes on to consider such things as searching for more information. He could of course treat it in the first place as a three-way decision, but this would complicate the matter for him even further.

(29) Matza, D., Delinquency and Drift (New York, Wiley, 1964)Google Scholar.

(30) Strictly speaking this arises only if the insurance company pays into court Sociothe sum of money it has offered in com-pensation. I gather that this is normally done.

(31) This information was reported at a seminar organized by the Centre for Sociothe Legal Studies, Oxford, in 05 1974Google Scholar.

(32) All the quotations are from Simmons et al., op. cit.