Sex differentials in the earnings of Ph.D.s.

Using a survey of two cohorts of men and women who received Ph.D.s in the years 1958-63 and 1967-72, the authors test two hypotheses: (1) that the relatively lower earnings of highly educated women can be explained largely by their career interruptions and by their lesser willingness to accumulate human capital in anticipation of such interruptions, and (2) that the differential in earnings between men and women increases with age because of career interruptions and that the gap narrows once women reenter the labor force on a permanent basis The findings do not lend support to either of these hypotheses, leading the authors to reject the proposition that the lower rewards of women Ph.D.s are primarily caused by their own voluntary decisions.     

Confidence intervals for batch properties based on both sampling and analytical variance

Summary If different random samples are taken from a heterogeneous batch and each sample is analysed once or in replicate, the average result is a point estimation of the true but unknown mean of the batch property concerned. In this paper, formulae are given for the numbers of degrees of freedom, which are necessary for interval estimations based on both the sampling and the analytical variance. It is essential to distinguish between situations in which estimates of the variances are available and in which they are not.     

Collective agreements and the law in the Republic of Ireland: a case study in legal enforceability1

On a small scale, collective agreements are already legally binding in Ireland. A look at their operation and effectiveness gives some indication of what might happen if Conservative proposals become law in this country.     

In quest of useful models for educational systems

Analytical flow models are examined as a means for describing the performance of an educational system. Enrollment functions based on specified input functions are developed for a single grade for four special cases. A standard form for approximation is introduced and its use described. This leads to the proposition that complex systems can be fabricated of modules defined in terms of these standard approximations. Applications of this approach, and problems expected in its use, are discussed, including the question of quality of the educational process. The paper concludes with some brief comments on measurement and evaluation of performance.