Information system design alternatives have long puzzled and frustrated those interested in improving information system design. Design alternatives exist in all phases of the information system including measurement, collection, storage, processing, communication and display.' This study concentrates on a simplification of the general design problem of comparing one specified information system, an information structure, with an alternative. Review of the literature indicates that research methodology for this question is incomplete, particularly with respect to empirical verification of hypotheses, with respect to the development of empirical measures of the usefulness of changes in the information system, and with respect to inclusion of the real decision maker into the analysis. Thus, a main purpose of this study is to develop and implement empirical means of measuring and comparing the effects of changing the design of an information system. An experiment is designed for this purpose. The experiment is a business game that requires the businessmen and student participants to reach decisions concerning production quantities, advertising purchases, and input mix. The independent (controlled) variable in the experiment is a difference in time delay before information is communicated.

This article presents information on an experimental study concerning the impact of budget feedback information and the budgetary process on decision and performance. Two types of learning which can be helped by accounting feedback are, learning of the system or model that affects payoffs and learning which decision strategies or policies are effective in coping with ill-structured and changing decision environments. Information useful in these two senses has previously been labeled model and action effectiveness values of information. In order to generate some empirical data on the value of such information, a set of business-game experiments was conducted. These experiments are described in the following section of this paper, but first consideration of sequential decision processes is necessary. The explicit hypotheses as to how each information structure and each motivational treatment are expected to affect performance and learning depend upon which theory of decision one bases his hypotheses on and which of several problem or environmental classifications the particular decision problem fits most closely.

The article presents the author's response to a commentary related to his article on the concepts of information value and accounting. The author said that he does not share the faith and enthusiasm of the the writer of the commentary as to the viability and practical utility of the pure information economics approach as an aid to choosing alternative accounting systems. He said that the commentary's writer seems to misinterpret the thrust of his argument. He also does not share the disdain of the commentary's writer concerning historical measures used as estimates of information value.

This article deals with the dual objective of presenting a framework in which research can be evaluated and which will stimulate and improve accounting research from a pedagogical stand-point. The framework developed views research as a decision problem and thus touches upon many of the concepts of decision theory. A methodological decision tree is presented as a useful technique for the planning and controlling of one's research strategy. Before the methodological decision tree approach can be adequately presented, several conceptual distinctions should be made, including a distinction between research methodology and research technique. It will be useful to consider methodology as the meta-decision process underlying selection of a particular research design. Whereas by research techniques, the notion of those analytical and technical tools available for research implementation is suggested. Consistent with these definitions "scientific method" is at the methodology level whereas statistical techniques such as regression and factor analysis are at the technical level.

This article develops and applies the concepts of model, action effectiveness and decision value of information in a production decision example. It is argued that a message has several types of potential value and that accounting information should be evaluated according to its learning characteristics in addition to its value in decision. The potential value of information in planning, in control, and in decision making empirically seems to exhibit itself in contributing directly to choice, in improving one's understanding of real-world relationships and in identifying payoff relevant actions. The need to focus on the potential values of information has arisen to counter the decision ethos or advocacy of many information system designers, to explain the role of feedback-oriented information such as many accounting systems, and also to explain several empirical and experimental observations. The concept of information value that derived from and focused upon statistical decision theory has been examined primarily from the standpoint of learning and has been extended to conform to what is argued to be the role of much accounting information. The design of accounting information systems can be expected to benefit to the extent that such concepts improve both our understanding and our methods of estimating the value of information.

Reviews the book "Systems Organizations, Analysis, Management: A Book of Readings," by David I. Cleland and William R. King.

Reviews the book "Systems Analysis and Project Management," by David I. Cleland and William R. King.