Finite sample distributions of statistics in change point analysis

This paper is concerned with derivation of finite sampling distributions of some statistics which appear frequently in change point analysis. The exact distribution of cusum test statistic is approximated by two methods. Approximations are presented and their accuracies are measured. We first consider the change point in mean problem and we study the exact distribution of change point estimator. Finally, we consider the change point in variance case.     

Social Capital in Association with Socioeconomic Variables in Iran

The objective of this study was to determine the various dimensions of social capital in Tehran and its association with socioeconomic variables. Two thousand, four hundred and eighty-four persons participated in the study through multi-stage stratified sampling in Tehran. The study questions were extracted from the World Bank integrated questionnaire. While respondents obtained over 50 % of the attainable scores in the ‘trust’, ‘cooperation’, ‘social cohesion’, and ‘political action’ dimensions, they obtained only 18.9 % of the attainable scores in the ‘networks’ dimension. The respondents in age groups older than 45 had higher trust scores than respondents in the 18–25 age group. Men had higher ‘network’ and ‘social cohesion’ scores than women (p ≤ 0.001). Respondents in the poorest wealth quintile had lower ‘networks’ and ‘social cohesion’ scores than the richest group and ‘political action’ scores higher than the richest group (p = 0.038). The lowest score in ‘Tehranians’ social capital was in the groups and networks dimension. Therefore designing effective interventions for such activities, particularly in women’s groups, is a priority.     

Integration of process planning and production planning and control in cellular manufacturing

Organisations willing to succeed in global competition have to integrate their internal and external processes. This especially includes planning and production control (PPC) processes. Optimised allocation of the production resources and quick response to demand changes result in lower cost and improvement of production performance. Practitioners and researchers have been trying to achieve these goals using production planning techniques. Although the results are significant, it seems necessary to integrate production operations in order to improve the production performance. The goals, information and decisions taken in production planning and control and process planning are often very different and difficult to integrate in Cellular Manufacturing (CM) environments. Designing an efficient PPC system and integrating it with process planning in a cellular environment is of the same importance. The following paper proposes first a comprehensive framework of integrated process planning and production planning and control in CM. Then, with respect to this framework and utilising the domain knowledge in the area of CM systems, an integrated model based on Integrated Definition Modeling Language is developed. The application of the models has been considered as a case study for a production system in electronics and telecommunication sector in a plant in Iran. The validity and completeness of the proposed model is tested by a panel of experts in the areas of production planning and control in CM environments.     

Modernization process and fertility change in pre- and post-Islamic Revolution of Iran

The purpose of this paper is to examine some of the possible causal links between modernization forces and fertility patterns of the Iranian provinces during three time periods, 1966, 1976 and 1986. A modernization scale was constructed using Iranian census data. Six indicators of industrialization, urbanization and universal education were used to develop the scale. The ratio of children under 5 years per women 15 to 44 years old was used as a measure of fertility. The findings show that modernization has proceeded upward in an almost consistent pattern in all the provinces during 1966–1986. The Islamic Revolution and Iran-Iraq War not only did not disrupt the modernization trend, it seems that both events accelerated the rate of change. The modernization indicators, individually and collectively, were significantly and inversely correlated with fertility ratios. However, the fertility ratios of the provinces substantially increased in the decade of 1976 to 1986. Our thesis is that the elimination of the national family planning program which happened in the early part of the post-Islamic Revolution had significant effect on the fertility increase of the period 1976 to 1986. The current active family planning program of the Islamic Republic of Iran suggests that the fertility rate of Iran, very likely, will decline in the near future if the current modernization trend and fertility regulation policy continue.An earlier version of this paper was presented at the 22nd General Conference of the International Union for the Scientific Study of Population, Montreal, Canada, 24 August 1993.     

Duality of weakly disposable technology

The axiom of weak disposability is frequently imposed in data envelopment analysis (DEA) models involving undesirable outputs such as pollution. This paper sheds new light on the economic interpretation of weak disposability by developing dual formulations of the weakly disposable DEA technology. We find that the economic implications of weak disposability on the multiplier DEA problem are two-fold: (1) the shadow prices of the undesirable outputs can be positive or negative, and (2) the economic loss of the benchmark cannot exceed the suck cost of the inputs. We interpret the second implications as a limited liability condition. The dual formulations developed in this paper enable one to estimate shadow prices of the undesirable outputs using the weakly disposable technology. The insights gained are illustrated by a numerical example and an empirical application to the US power plants.     

Nonparametric change point detection for periodic time series

We consider detection of multiple changes in the distribution of periodic and autocorrelated data with known period. To account for periodicity we transform the sequence of vector observations by arranging them in matrices and thereby producing a sequence of independently and identically distributed matrix observations. We propose methods of testing the equality of matrix distributions and present methods that can be applied to matrix observations using the E-divisive algorithm. We show that periodicity and autocorrelation degrade existing change detection methods because they blur the changes that these procedures aim to discover. Methods that ignore the periodicity have low power to detect changes in the mean and the variance of periodic time series when the periodic effects overwhelm the true changes, while the proposed methods detect such changes with high power. We illustrate the proposed methods by detecting changes in the water quality of Lake Kasumigaura in Japan. The Canadian Journal of Statistics 48: 518–534; 2020 © 2020 Statistical Society of Canada     

Vertical Mosaic among the Elites: The New Imagery Revisited*

La parution de Vertical Mosaic (1965) par J. Porter, Canadian Corporate Elite (1975) par W, Clement et State Elite (1980) par D. Olsen a provoqué dans les milieux universitaires un vif débat sur la surreprésentation des ethnies britanniques et la sous-représentation des ethnies française et “autres” au sein des élites canadiennes. Cependant, dans une série d'articles récemment parus, R. Ogmundson et J. McLaughlin ont mis en doute la thèse très répandue selon laquelle il existe dans la société canadienne un lien étroit entre l'origine ethnique et 1'appartenance à une élite. L'auteur de cet article examine les données avancées par Ogmundson et McLaughlin (1992) et démontre, en faisant une corrélation avec la composition ethnique de la population canadienne, que leur nouvelle interpretation peut induire en erreur. L'analyse fait valoir en effet une predominance stable des ethnies britanniques de 1935 à 1987 et ce, pour chaque catégorie des élites. Elle met aussi en évidence le fait que, conformément aux théses de Porter, Clement et Olsen, les ethnies française et “autres” ont augmenté leur représentation au sein des elites, mais n'ont pas atteint un niveau de representation proportionnel à leur nombre. The over-representation of British and the under-representation of French and “other” ethnic groups amongst the Canadian elites has been a part of the Canadian academic consciousness since the publication of Porter's Vertical Mosaic (1965), Clement's Canadian Corporate Elite (1975) and Olsen's State Elite (1980). Recently, a series of articles by Ogmundson and Ogmundson/McLaughlin has raised doubts about the vertical ethnic distribution of power among elites, suggesting that conventional wisdom should be abandoned. This paper uses the data provided by Ogmundson and McLaughlin (1992), standardizes them to the ethnic composition of the Canadian population, and shows that the new imagery provided by these authors is misleading. The analysis demonstrates a stable pattern of British dominance from 1935 to 1987 for each category of elites. It also reveals that, consistent with Porter, Clement and Olsen, the French and “other” ethnic groups have improved their participation at the elite levels but remain under-represented in relation to their populations.     

Structurally consistent probabilities of selecting answers

This paper offers a procedure for specifying probabilities for students to select answers on a multiple-choice test that, unlike previous procedures, satisfies all three of the following structural consistency conditions: (1) for any student, the sum over questions of the probabilities that the student will use the correct answers is the student's score on the test; (2) for any student, the sum over possible answers of the probabilities of using the answers is 1.0; and (3) for any answer to any question, the sum over students of the probabilities of using that answer is the number of students who used that answer. When applied to an exam, these fully consistent probabilities had the same power to identify cheaters as the probabilities proposed by Wesolowsky, and noticeably better power than the probabilities suggested by Frary et al.