Search results

It has often been said that a great part of the strength of Aslib lies in the fact that it brings together those whose experience has been gained in many widely differing fields but who have a common interest in the means by which information may be collected and disseminated to the greatest advantage. Lists of its members have, therefore, a more than ordinary value since they present, in miniature, a cross‐section of institutions and individuals who share this special interest.

Statistical thinking is an intrinsic part of the quality movement. Helped by initiatives such as Six Sigma, there is greater acceptance of the importance of data analysis and a general trend towards embracing numeracy. It is timely to review the emergence of statistical thinking and consider the good and bad features resulting from its application in a wide range of sectors.

The paper first defines statistical thinking and justifies its importance to the quality movement. The achievements from the past 25 years are then considered sector by sector along with their collateral damage.

The following lessons are proposed for the next 25 years: statistical thinking needs to expand its remit to include more aspects of analytical thinking becoming what may be called wider statistical thinking; statistical thinkers have ground-breaking ideas and need to communicate with managers at the top of the hierarchy to ensure that both the thinkers and the ideas have the influence they deserve; whilst learning from past successes, the quality movement must be mindful of knock-on effects and nurse a holistic viewpoint; expect the unexpected.

Statistical thinking is gaining more prominence in all sectors and is used within the quality movement to make major progress as well as major upsets. It is important that the quality movement treads carefully and makes sure that society as a whole benefits from the ever increasing drive for improvement.

Aarhus Kommunes Biblioteker (Teknisk Bibliotek), Ingerslevs Plads 7, Aarhus, Denmark. Representative: V. NEDERGAARD PEDERSEN (Librarian).

This chapter enquires into the contribution of two British writers, Herbert Somerton Foxwell and Henry Riverdale Grenfell, who elaborated upon the hints provided by Jevons towards a description of long waves in the oscillations of prices. Writing two decades after Jevons, they witnessed the era of high prices turning into the great depression of the last quarter of the nineteenth century, the causes of which they saw in the end of bimetallism. Not only did they take up Jevons’s specific explanation of the long fluctuations, but they also based their discussion upon graphical representation of data and incorporated in their treatment a specific trait (the superposition principle) of the ‘waves’ metaphor emphasized by the Manchester statisticians in the 1850s and 1860s. Their contribution is also interesting for their understanding of crises versus depressions at the time of the emergence of the interpretation of oscillations as a cycle, which they have only partially grasped – as distinct from the approach of later long wave theorists.

That part of human behaviour which is not rigidly determined by external constraints can be considered as a sequence of more or less free choices. One can talk of a choice only if it is one of several alternatives, and of a free choice only in so far as it is not determined by conditions over which the individuals have no control. Moreover, we recognise an alternative only if it is actually elected by at least a fraction of a population. This leads to the concept of statistical freedom. Postulates are formulated which must be satisfied by any numerical measures of statistical freedom, and certain mathematical expressions are proposed which are shown to conform to these postulates. Statistical freedom has two fundamental features, which appear as factors in its numerical measure: diversity and independence. The measures of diversity and independence are derived in the first place front a certain model of society, but once they are obtained, the model is discarded, and the statistical coefficients are justified by their mathematical properties. Safeguards against arbitrary manipulation of statistical material are discussed, and the potential use of the new measures is illustrated by application to the problem of choice of profession.

1. Documentation, library and book organizations (0) American Documentation Institute

This study was inspired by the concept of entropy as applied to the measurement of the extent of information in the communication theory of Wiener and Shannon. Like its precursor, entropy in statistical mechanics, it is a convenient measure of the uncertainty or unpredictability of a system or of a process containing an element of contingency. By keeping in mind the fact that in human circles unpredictability is often allowed to pass under the sacred name of liberty, it is possible to perceive an opening for mathematics into a domain which was until now inaccessible to the mathematician. It was four years ago that we made a first attempt to enter that field in our “Essay on the Mathematical Theory of Freedom”, presented to the Royal Statistical Society of London.

Adam Smith, it is generally acknowledged, founded the modern discipline of political economy with the study entitled An Inquiry into The Wealth of Nations (1776) which he built upon the ethical system he presumed to exist in his Theory of Moral Sentiments (1759). Ricardo took Smith's observations somewhat further with his publication of On the Principles of Political Economy and Taxation (1817). When John Stuart Mill wrote his Principles of Political Economy in 1848, his considerations of economic processes were intimately connected with the political. By the time Marx published Das Kapital as a critique of political economy in 1867 the term was entrenched in both academic life and in common parlance and political circles. The study of economics was an integral part of the study of the state. Ironically, however, political economy was about to be upstaged by the development of economics as a separate and positivist discipline. William Stanley Jevons had published his “Brief Account of a General Mathematical Theory of Political Economy” in the Journal of the Royal Statistical Society in the previous year. This was much more widely read at the time than Das Kapital. By 1890, Alfred Marshall had published his Principles of Economics. The book began with these words: “Political economy or economics is a study of mankind in the ordinary business of life.” The great tradition of seeing economics as an integral part of politics and vice versa was disappearing. However, though economists were anxious to convert that part of political economy known as economics and see it as a scientific discipline, the reality, that is the integrated nature of the state and the economy, remained. Simply because certain ideologues decided to separate politics from economics did not mean that the state in any sense disentangled itself from the economy or the economy from the state.

This chapter provides a presentation about Chapter 1 of The Balance of the National Economy, 1923–24, edited by Pavel Illich Popov. The Balance was issued in June 1926 by the Central Statistical Administration (CSU or TsSU) of the USSR, which Popov had headed from July 1918 to January 1926. In the first part of our chapter, we show how Popov’s work on the balance of the national economy was rooted in the specific scientific and political culture of zemstvo statisticians inherited from the Tsar. Statistical inquiry was considered an objective scientific process based on international standards. Furthermore, like zemstvo statisticians, CSU statisticians developed great autonomous political power. The balance of the national economy was built according to these principles, which met harsh criticism from revolutionaries and Bolsheviks. In the second part, we analyze the contents of Popov’s Chapter 1, especially the theoretical foundations of the balance and its connection with Soviet planning. In the third part, we discuss the balance’s significance in the years 1926–1929, years which ended the NEP and launched the first Five-Year Plan, so as to understand how CSU’s balance didn’t become a standard Soviet statistical instrument and was discarded as a “bourgeois” device.

The role of uncertainty in manpower planning is discussed and illustrated by reference to manpower planning in the Civil Service. The nature of the planning process is described, and an account given of the principal statistical methods employed. Uncertainty is classified in terms of random error, specification error, estimation error and data error. Strategies for containing error are discussed, and examples used to illustrate the relative importance of each type. Parameter specification error is usually dominant, and experience suggests that sensitivity analysis is the most practical way of dealing with it. The mutually‐dependent roles of statistician and manager in the process of working towards effective decisions, and the importance of a flexible and continuous planning process which can adapt to the unexpected, are both stressed.