Analysing and aiding decision processes

Analysing and aiding decision processes

Book Reviews 252 at a glance where the example ends and the text continues. Otherwise though the layout of the book is good, the material being well...

282KB Sizes 3 Downloads 101 Views

Book Reviews

252

at a glance where the example ends and the text continues. Otherwise though the layout of the book is good, the material being well-spaced and very clear. In summary, this is a book on well-established conventional undergraduate material, nicely presented and expla;,ned, and one wtfich is certainly to be recommended.

Valerie ISHA M Department of Statistical Science University College London London, United Kingdom

John GRIEVE SMITH (ed.) Strategic Plarming in Nationalised lndustric~ Macmillan, London, 1984, xvi + 268 pages, £25.00 This book is based on papers given at a conference held in Cambridge, England in September 1983. The nationalised industries referred to in the title are the (then) British state-owned industries. Although most papers ma.r~e more general theoretical points, all the practical examples discussed draw on this British background. The book starts with two scene-setting cimpters: Taylor reviews the general nature of stratef, ic planning and Grieve Smith relates these general consi,Jerations to the particular case of state-owned industries. l'here follow five papers on the question of how state or public control of nationalised industries has been exercised and how it might be. These papers, among other things, serve to remind us of the range of control measures that have been tried, and how variable their success has been. The papers also draw attention to some control problems that have barely been faced, let alone resolved. Various interesting suggestions are made for improving control. Next come thl'ee papers on methods and techniques, which review price-setting methods, forecasting and the (encouragingly high for OR people) use made of models..'I~,e two final chapters are historical critiques of strategic planning and its outcome in two specific cases: the transport sector and the coal ;rod electricity part of the energy sector. An editorial post-script, drawing attention to current problems and ke~ issues to be resolved, closes the book. The individual papers are generally well writ-

ten, well researched and thorough within themselves. Inevitably, as in all books based on conference proceedings, there are papers or parts of papers with which those with prior interest in the subject matter will find they are already familiar, and others were those approaching the subject for the first time will wish for fuller background. On the whole, however, the papers steer a useful middle course. The book as a whole provides a pretty though review of its subject matter. An obvious missing component is contributions direct from the principal actors: the politicians, those managing the industries, those working in them, and consumer representatives. ,(The great majority of the contributors are academics, and the book has a sense of commentary on the actions and motives of others as opposed t,o the tanginess of good autobiography). Nor is there a contribution, which would certainly have been relevant to discussion of the energy sector, from a committed environmentalist. But one cannot have everything. At £25, the book is priced beyond the means of most individual buyers. It should, however, find i,'s way on to the bookshelves of those libraries serving people with an interest in strate'.3ic planning, state-owned industries and the problems of public-sector control generally.

G.H. MITCHELL National Coal Board Harrow, United ffingdom

P. H U M P H R E Y S , O. SVENSON and A. VARI (eds.)

Analysing and Aiding Decision Pro~esses Volume 14 in: Advances ~'n Psychology NorthHolland, Amsterdam, 1983 + 565 pages, Dti.135.0C The first flmr or five sections of this book p,esent an ediled selection of papers presented at the Eighth R,;search Conference on Subjective Probability, Utility and Decision Making, held in Budapest in August 1981. The fifth section presents papers given at a symposium nn the validity of studies on heuristics and biases at this conference. They are not edited and they are intended to stimulate discussion among a wider audience. This book p~resen~s a wide vzriety of contribu-

Book Reviews

tions of an interdisciplinarj nature and it must have been very difficult for the editors to design papers to categories. Those categories or sections of the book.are therefore not at all disjunct, but their topics express the main areas of concern of the Budapest conference. The titles of the first four sections are: "Societal Decision Making", "Organisational Decision Making", "Aidir~g the Structuring of Small Scale Decision Problems" and "Tracing Decision Processes". Throughout the book the emphasis is on decision processes and structure and their applications rather than on formal modelling in isolation. The papers of se:tion I. "Societal Decision Making", are not as far removed from operations research as one could probably expect. In particular the paper by Lathrop and Linneroth: "The Role of Risk Assessment in a Political Decision Process" gives rather detailed and interesting descriptions of a case in political decision making which can be very useful for OR analysts. It is particularly interesting that the amhors distinguish between different kinds of risk and evidence which might lead to a more appropriate modelling of uncertainty. Section I contains three more case study type papers which could certainly be used at least to analyse them critically in order to find more potential applications of OR approaches. Section II, "Organisational Decision Making", is of a more philosophical nature. No empirical results are presented, but rather general discussions on pitfalls of decision analysis and on its relationship to other areas such as systems analysis, organisational development and similar areas. Section III, "Aiding the Structuring of Small Scale Decision Problems", is a very heterogeneous and probably therefore particularly interesting section. Its scope runs from human engineering via fuzzy structural modelling to multi-attribute utility theory. The contributions in that section are partly of a more descriptive character; they report on empirical experimentation (non-expert use of a computer based decision aid) or they are more of a modelling type (fuzzy structural modelling). Section IV, "Tracing Decision Processes", opens with the theoretical and stimulating paper by Montgomery who argues that the decision process may l:,e characterized by search for good arguments supporting the later chosen alternative. Thus the justification of the choices is seen as the ulti-

253

mate goal of a decision process. Montgomery suggests one possible way to reach this ~oal: That of finding or creating a dominant situation favouring only one choice alternative. This theme is taken up again in a paper by Ranyard and Crozier who study biddings on individual gambles, choices between two gambles, and decisions with three or five alternatives. The verbal reasons collected in these situations were classified in different categories and different heuristic approaches are described and investigated. Two papers are concerned with the use of verbal protocols for empirical research. They will be more of interest to scientists doing empirical reseach in decision processes. The rest of the papers in this section concentrate on the analysis of information processing as decision making processes. Klayman analyses information search patterns and shows the difficulties in relating these to decision rules because the search characteristics of a given decision rule or a sequence of decision rules may be variable and task dependent. The papers by Goldsmith, Salin and by Smith, Ferrell examine the aspects of subjective probability structures and reflect on the differences between subjective probabilities, second order probabilities, confidence statements and so on. In summary all papers of this .,;ection show the importance of studying decision making as a process and the value of data from process tracing studies for improving our knowledge about human decision making. It was mentioned that the contributions in this book are very heterogeneous in character and that it must have been difficult to structure them by assigning them to five sections of this book. To facilitate the study of the book the editors provide for each section a very good introduction which shows the inner relationship of the papers contained in the different sections. So the introductions are very helpful when working through the book and one should actually start the study of this volume by reading all of those introductions first. Even though the volume is published as volume 14 of the series "Advances in Psychology" it contains many pieces of important knowledge from the point of view of operational research. This is particularly true for the Sections I, II and IV. The book can therefore be recommended to operations researchers who are looking around for rele',ant

Book Reviews

25,,

ideas beyond the scope of classical OR methods such as mathematical programming, queuing theory etc.

H.-J. Z I M M E R M A N N R WTH Aachen Aachen, Germany, Fed. Rep.

Jrzef KOZIELECKI Psychological Decision Theory Reidel, Dordrecht, Dfl.125.00

1982,

xvi + 4~.3 pages,

This book is a revision of the Polish originzl published in 1975. The bibliography has been extended to 1976 and 1977 and there is one referem-e dated 1978. Not having access to the originM edition, the reviewer is unable to say what other changes were involved in revision. The book gives a very thorough and systemati treatment of decision theory, through a discussion of the decision maker and decision tasks, representation, utility and subjective probability, tc algorithmic and heuristic strategies of risky choice, dynamic risky tasks and open decision tasks. Th,~ later chapters deal with the environment of the :leeision task, personality and risky decisions, group decision making and applications of ps~cchoogical decision theory. The discussion is clear and enlivened by good t~xamples and illustrations from real-life and published research. There is nothing awkward about the translation, which runs smoothly and yet with a crisp firmness of statement. The bcok obviottsly does not provide an account o ~"the 'state of the art' in decision studies. It is an e~:positi,3n "addressed to psychologists, sociologists, management theorists and economists concerned with decisions made by real persons" and "may also prove usefu 2as a text-book for graduates and, in part, for undergraduates in psychology". The 1eviewer has tried it out as the central text of a course on decision theory for final-year undegraduate psychology students. Their reat, tion was ver3.., favourable and it was felt that it had only tw,~, mirior shortcomings in this context. The first is ~hat the book begins w~th what is, in essence, the volt Net~mann and M:~rgenstern paradigm. While there are many, and velcome, subsequent refer-

ences to findings in areas of psychology other than decision making in the 1950's and 1960's, the book lacks an introduction dealing with the origins of decision theory. It is true that decision theory grew rapidly after the publication of von Neumann and M~rgenstern's book in 1944, but some account of its :nots in philosophy, statistics, economics and, to a lesser extent, in psychology of the 1920's and 1930's would give the reader an appreciation of ,e conceptual systems out of which this paradigm arose. The second shortcoming is the obvious one '..~t the later chapters of the book need to be supplemented by references to more recent published work. But these are small disadvar:tages in using what is, in general, a very sound text. The book is highly recommended to all those for whom it was apparently intended. One can only hope that the publishers will feel disposed to produce a soft-back edition more in the price-range appropriate to student use.

M.F. MOOP,E University of Stirlvrg Stirring, United Kingdom

Matthew B. MILES and A. Michael HUBE];:MAN Qualitative Data Analysis: A Sourcebook of New Methods Sage, Beverly Hills, 1984, 263 pages, £27.50 Very often, qualitative research methods are regarded and presented as the fullfledged alternative for its reputed quantitative counterpart. According to this widespread opinion, qualitative methods are those that appropriately handle the object and problems of the social sciences and, as a fringe but welcome benefit, are much easier to practise than quantitative methods. When this is true, however, it is true as z consequence of the lack of explicit methodological conventions and procedures in qualitative re-. search. In discussions, qualitatiw~ researchers ofter even demonstratively seem to consider this lack as the hard core of their methodological principles. This book was written " t o address a critical need faced by researchers in al~ fields of the human sciences (-): How can we d~aw valid meaning