Search

This article comments on George Zito's article "A Note on the Population of Seventeenth Century London," published in the 1972 issue of the journal "Demography." This article shows his concern in hypothesis of linear population growth for London and whether he has provided sufficient statistical evidence to warrant his conclusions. Even if the resulting t-value were large enough to lead to a rejection of the hypothesis, one should not discard his equation because this equation might be the best estimate of the underlying relationship up to 1690, whereas some other model of London's population growth might pertain to the period from 1690 to 1801. By judging the suitability of different equations, on the basis of how close they come to the 1801 census value, Zito seems to be assuming that the linear model describing London's population growth is appropriate throughout the entire 250-year period from 1550 to 1801. This t-value is not strong enough to reject a one-tailed test of the hypothesis at the five percent level.

The effects of education on current migration propensities in the United States are examined for each age group, and an estimate is made of how these age-specific differentials cumulate over a lifetime in determining the number of times that individuals at different educational levels can expect to move over the course of their lives. The independent effects of age, education, and occupation are also examined. Two other sections investigate trends in educational selectivity of migration in the United States and international comparisons of occupational differences in migration. [ABSTRACT FROM AUTHOR]

This article focuses on the two multivariate analyses of criminal statistics in England and Wales, since the war. The numbers or the percentage of the population in a given age-group have been used as predictor variables, even though different birth year cohorts are involved at each observational point. From an examination of the data on recorded delinquents for England and Wales for 8-20 years-olds from 1946-57, the author concluded that children who passed through their fourth and fifth years during war-time were more likely to be delinquent. Author have normally used data about age-specific delinquent rates in different years; but cannot immediately infer the number of identified delinquents in a particular birth year cohort, from the delinquent rates observed in the succeeding years. Some people have therefore suggested that the only way to gauge the number of delinquents of any given birth-group is to find out how many individuals of that birth-group have become delinquent before, say, adulthood, or before retirement. This means following up the careers of a given cohort over a long period and this is not always possible.