A NOTE ON A MULTIVARIATE GENERALIZATION OF THE KRUSKAL-WALLIS TEST

Marketers are often interested in testing whether the mean vectors of multivariate distributions are equal. The test usually applied, one-way MANOVA, assumes the distributions are multinormal. Unfortunately, this assumption is not supported in many studies. As an alternative, a nonparametric multivariate one-way analysis of variance procedure is presented.     

The Predicament of Population Aging: A Review Essay*

Ken Dychtwald, Age Power: How the 21st Century Will Be Ruled by the New Old Richard Leete (ed.), Dynamics of Values in Fertility Change Donald T. Critchlow, Intended Consequences: Birth Control, Abortion, and the Federal Government in Modern America Bernard Jeune and James W. Vaupel (eds.), Validation of Exceptional Longevity     

A note on the estimation of selective values from recaptures of marked animals when selection pressures remain constant over time

Summary An equation is given for the estimation of selective values from data obtained by mark-recapture experiments, assuming that selective pressures remain constant while the experiments are carried out. The equation does not have an explicit solution but can readily be solved using a trial-and-error method. The use of the equation is illustrated on some data reported byKettlewell et al. (1969) from an experiment involvingtypica andedda morphs of the mothAmathes glareosa. It is found that theedda morph apparently had a selective advantage of about 12% per day compared to thetypica morph and that this is significantly different from zero. Using another methodKettlewell concluded that the selective advantage of theedda morph was only 7% and that this was not statistically significant.     

Men matter: Additive and interactive gendered preferences and reproductive behavior in kenya

The extent of men’s roles in reproductive decision-making in Africa is a subject of contention. Despite the volume of work on the roles men play in fertility decisions, there have been few attempts to derive direct empirical estimates of the effect of men’s preferences on reproductive behavior. I employ 1989 and 1993 Kenya Demographic and Health Surveys to examine the relative roles of the reproductive preferences of males and females on contraceptive use. Additive and interactive measures of preferences document a significant effect of men’s preferences, which may eclipse women’s preferences. The implications of these findings are discussed.     

Effects of Group Practice of the Transcendental Meditation Program on Preventing Violent Crime in Washington,D.C.: Results of the National Demonstration Project,June--July 1993

This paper reports the results of a prospective experiment in which a group of approximately 4,000 participants in the Transcendental Meditation and TM-Sidhi programs of Maharishi Mahesh Yogi assembled in Washington, D.C., from June 7 to July 30, 1993. It was hypothesized that levels of violent crime in the District of Columbia would fall substantially during the Demonstration Project, as a result of the group's effect of increasing coherence and reducing stress in the collective consciousness of the District. A 27-member Project Review Board comprising independent scientists and leading citizens approved the research protocol and monitored the research process. Weekly crime data was derived from database records provided by the District of Columbia Metropolitan Police Department (DCMPD), which are used in the FBI Uniform Crime Reports. Statistical analysis considered the effect of weather variables, daylight, historical crime trends and annual patterns in the District of Columbia, as well as trends in neighboring cities. Consistent with previous research, levels of homicides, rapes and assaults (HRA crimes) correlated with average weekly temperature. Robberies approximately followed an annually recurring cycle. Time series analysis of 1993 data, controlling for temperature, showed that HRA crimes dropped significantly during the Demonstration Project, corresponding with increases in the size of the group; the maximum decrease was 23.3% (p < 2 × 10–9) [24.6% using a longer baseline, with 1988--1993 data (p < 3 × 10–5)], coincident with the peak number of participants in the group during the final week of the assembly. When the same period in each of the five previous years was examined, no significant decreases in HRA crimes were found. Robberies did not decrease significantly. However, a model that jointly estimated the effect of the Demonstration Project on both HRA crimes and robberies showed a significant reduction in violent crimes overall of 15.6% (p = 0.0008). Further analysis showed that the effect of the coherence-creating group on reducing HRA crimes could not be accounted for by additional police staffing. The time series analysis for HRA crimes gave results that are highly robust to alternative model specifications, and showed that the effect of the group size was cumulative and persisted after the Demonstration Project ended. Also, calculation of the steady state gain based on the time series model predicted that a permanent group of 4,000 coherence-creating experts in the District would have a long-term effect of reducing HRA crimes by 48%.     

THE EFFECT OF RETAIL PROMOTIONAL ACTIVITIES ON SALES*

This article reports the formulation and results of a multiple equation econometric model to relate retail sales by brand and package size to retail promotional variables for a branded, frequently purchased grocery product. Primary emphasis is placed on the formulation of the model as opposed to the results obtained from its use.