Measuring the mortality impact of health interventions is not straightforward; there is now ample evidence of this. So why bother? Clearly, the utility, sensibility and feasibility of measuring the mortality impact of health interventions depend on many factors, which makes it difficult to offer simple prescriptions. The failure of those demanding measurement to appreciate this has led, on the one hand, to the proliferation of largely unsubstantiated claims on the achievements of health interventions and, on the other hand, to disillusionment about the prospects of demonstrating any impact on mortality. It is not difficult to account for the complexity of this area nor the current state of uncertainty. Progress has largely reflected developments in its three components—mortality measurement, health research and programme evaluation. It is thus not surprising to find the almost exclusive emphasis on measuring impact on infant and child mortality; in developing countries, child health is given a high priority in the allocation of resources by governments and donor agencies, and the techniques for estimating child mortality are generally more advanced.

Humans in many societies are known to mate, or marry, assortatively for a number of characters such as eye colour, height, IQ and place of birth. In this assortment an element of active choice may be involved. It is not known whether this choice is genetic. Two models of human mate choice are examined in which both males and females can express a mating preference. In the first, ‘sexual’ preferences can be expressed for any phenotype not necessarily one's own; in the second, preferences are only expressed for an individual's own phenotype. The results of the examination indicate how much active choice would be needed to account for the observed correlations between human mates, and suggest whether human mating preferences are more likely to be sexual or assorting.

As part of an ongoing study on young adult psychological and social development, data were obtained through parental reports on the present residences and educational and occupational attainments of 718 present or former residents of Hawaii (average age 31 years). These subjects, as well as their parents, had been tested between 1972 and 1976 on measures of cognitive abilities and personality. The extent of emigration to the mainland in this middle to upper-middle class sample was over 40%. On average, former Hawaii residents now living on the US mainland were of higher intelligence and educational background than their counterparts still living in Hawaii. Differences were also found for number of children, crossethnic marriages, and occupational attainment (males only). In addition, parents of US mainland residents scored significantly higher on measures of cognitive abilities and education than parents of current Hawaii residents.

Birth history data from women in the 1975–76 Bangladesh Fertility Survey were used to search for intentions to replace dead children. The median intervals between successive births of orders (i) and (i + 1) were not shorter when some siblings of orders below (i) had died. Nor was the median duration between the death of a child and the first posthumous birth shorter when the dead child was a boy or when it was survived by fewer than two brothers. The median intervals were generally shorter when the mother lived in an urban rather than a rural area but this difference was attributable only to the shorter duration of breast-feeding by urban women. These results disputed the notions that the timing of births was deliberately quicker to replace a dead child, that attempts at replacement were sex-selective, or that child replacement intentions were stronger in urban than in rural populations.

Several new concepts are used to describe contraceptive use histories for nearly 1200 women in Peninsular Malaysia. These histories are summarized by 81 episode histories. Transition matrices provide useful summaries of the changes women make in their contraceptive practice from one pregnancy interval to the next. Data from the mid-1940s to mid-1970s, during which period there was a dramatic increase in contraceptive use, reveal considerable inertia in individual couples' contraceptive practice. Persistence with a method was greater the less effective the method: while 86% of couples using no method in one interval used no method in the next, only 56% of couples using the pill in one interval also used it in the next. Virtually all transitions are of three types: continuation with the same method, a change from no method to some method, or a change from some method to no method. For only 1% of all pregnancies did couples use one contraceptive method before a pregnancy and a different method after the pregnancy. Differences are examined by calendar year and education.

The randomized controlled trial has become the standard basis for the evaluation of new therapeutic agents and procedures (and for measuring the protective efficacy of new vaccines or for assessing the value of screening procedures). Patients, who have met the criteria for eligibility and have agreed to participate in the trial, are allocated on a random basis to the alternative therapies under consideration. In order to avoid possible bias in the handling or assessment of these groups, a double blind procedure is preferred; the therapy given is not known to those who administer it, to those who assess the course of the disease thereafter, nor to the patients themselves. There is an extensive literature on clinical trials covering their logic and history, modern developments and the many complex, often controversial, issues that such trials have provoked. Not all issues have been fully resolved but by and large the principle, the practice and the ethical concerns of clinical trials are worked out and firmly established.

Using medical records from maternity clinics in the two Norwegian cities Oslo and Bergen, the effect of lactation on infant mortality during the period 1860–1930 is examined, comparing those who were and were not breastfed in a total of about 6900 live born infants. The mortality of children not breast-fed was nearly three times that of those who were breast fed. In a Cox regression analysis the infant's year of birth and the mother's marital status were found to influence mortality in addition to lactation. Children born to unmarried mothers experienced a mortality about twice that of those born to married mothers, both during on-going lactation and in the absence of lactation, up to about 1915. Children born in Oslo had a slightly higher mortality than those born in Bergen.

The duration of lactation was found to have a continuing protective effect on infant survival after weaning—the longer the duration, the lower the mortality after cessation of lactation.

Previous studies of the seasonality of pregnancy and birth have typically been based on monthly numbers or proportions of pregnancies or births (i.e. their distribution throughout the year) without any explicit accounting of the size of the population actually at risk of pregnancy by month. From data on monthly populations at risk and monthly probabilities of pregnancy this paper determines the relative contribution of each component to the monthly distribution of pregnancies. The data come from the National Survey of Family Growth, Cycle III. While there is some seasonality in the size of the at-risk population, it contributes little to the overall monthly variation of pregnancy, which is determined mainly by seasonality of pregnancy probability (fecundability).

In India an increase in female employment outside the home has occurred during the last few decades, especially in urban areas. A working woman may face difficulties in attempting to fulfil the demands of both worlds, at home and outside, while a housewife may feel tired and irritated with her household chores and financial dependence. All these may cause stress for these groups of women. The present study compares a group of working mothers with their non-working counterparts with respect to: (a) stress level, measured in terms of their anxiety score; and (b) certain general indicators of health including a broad measure of stress. The results show that anxiety and health scores of the two groups of women are similar. Further, the health score and anxiety score seem to be correlated, more clearly among the working mothers.

This study examines contraceptive method switching among married women in the US. It enquires first into the effect of method type and women's socioeconomic characteristics on the risk of switching to each of six methods, including no method, and secondly into the previous methods used by women who adopt specific contraceptive means. The results indicate a great deal of circulation among all method types and of movement to unprotected intercourse. The adoption of sterilization is greatest among women not previously using any contraceptive method.

This study examines the biological influence of infant death on subsequent fertility in three Asian countries—Indonesia, Nepal and Sri Lanka, comparing the birth interval between two consecutive births up to the sixth birth by survival status of the preceding infant among breast-feeding women not using contraception.

There is consistent evidence of biological influence in each of the three countries. Infant death shortens birth intervals by up to 30%, though its influence varies between the countries.

The relationship is examined of child mortality in Ghana with six socioeconomic factors: mother's type of place of residence, education, occupation and work status, and current husband's education and occupation. Using data from the 1979–80 Ghana Fertility Survey, Trussell's marriage duration model was employed to estimate probabilities of dying at exact ages 2 and 5 in different population subgroups. The two education variables (mother's and husband's education) have the largest effect on child mortality, followed by husband's occupation and mother's occupation, in that order. In order to reduce child mortality to tolerable levels, expansion is urgently required of the medical and health services, balanced by an equal development of education, particularly of girls, help being given especially to the rural areas where the majority of the population live.

Evidence has accumulated to suggest that the mean IQs of Orientals in the United States and in the countries of the Pacific Basin are higher than those of Whites (Caucasoids) in the United States and Britain. This paper presents evidence from IQ tests on 4858 6-year-old Chinese children in Hong Kong. On the Coloured Progressive Matrices these children obtained a mean IQ of 116. Samples from Australia, Czechoslovakia, Germany, Romania, the UK and the US obtain IQs in the range 95–102. It is suggested that these results pose difficulties for the environmentalist explanations commonly advanced to explain the low mean IQs obtained by some ethnic minorities in the United States.

When assessing the effects of a programme one must first consider its objectives. When applied to the improvement of health, the word programme implies an intervention which is sustained, complex, dealing with more than a single aspect of one disease and usually going beyond simple provision of fixed curative facilities. How a programme is assessed will vary with the different components and the underlying purpose of the assessment. It is important to ask just what information is needed and whether or not it needs to be collected on a longitudinal basis.

The seasonality of births in the period 1871–1977 is studied in a rural north-western Spanish population. Based on a total sample of 11,695 birth registrations, temporal variation is analysed. For siblings, according to family reconstitution, the total family size, the legitimacy of the child, and birth order are considered. A coefficient of birth month dispersion is defined and estimated for each family. Intra-family variation is related to inter-family coefficients in order to determine whether the local seasonal pattern of births may be partly explained by family characteristics.

Patterns of contraceptive method change and their association with reproductive motivation and contraceptive failure are examined using data from a 1986 survey of family planning behaviour of rural Sinhalese married women aged 15–44 in seventeen districts of Sri Lanka. A notable degree of rationality in contraceptive method changes occurs with family formation. The attempt to control unwanted fertility leads to more efficacious use of contraception, including traditional methods. Some methodological improvements to the analysis of contraceptive switching are indicated. Implications of the findings for programmatic emphasis on permanent versus non-permanent modern methods are discussed.

The Gambia provides an unusual opportunity for the application of cost analysis to health due to a relatively long history of immunization and recent empirical research. The results should apply not only to immunization programmes but also to a variety of types of primary health care in developing countries. In particular, well-based estimates of the cost per case averted and the cost per death prevented by alternative health interventions should prove to be of widespread interest and usefulness.

This paper investigates the association of early childhood mortality (between birth and second birthday) with ethnicity in Nepal, based on data from the 1976 Nepal Fertility Survey, which was part of the World Fertility Survey. The approach is through a series of hazard models, which incorporate ethnicity, year of birth, mother's illiteracy, father's illiteracy, rural-urban residence, region, sex, maternal age, survival of previous birth, previous birth interval, and breast-feeding as covariates. Ethnic differentials in early childhood mortality are not explained by the other socioeconomic and demographic covariates, except for a modest effect of illiteracy, but the remaining covariates explain a great deal of variability in early childhood mortality itself. Analysis using an improved specification of breast-feeding as an age-varying covariate indicates, on average, that breast-feeding, relative to not breast-feeding, reduces age-specific mortality risks during the first 2 years of life by 76%, a very large effect.

This paper presents fertility and child mortality estimates for agricultural households of the Gondar and Hararge regions, based on the 1981 Rural Demographic Survey. The study shows that fertility and child mortality are quite high in both regions as in the rest of the country. However, Hararge has significantly higher mean parity and child mortality than Gondar.

The influences of household economic condition, maternal education, sex, and nutritional status of children on mortality were examined using multivariate analytical techniques. Weights of around 1700 children aged 2–60 months in five villages of Matlab, Bangladesh, were taken during the first half of 1981. The children were followed for 18 months and their survival was recorded. The severely malnourished children had a risk of death nine times that of their counterparts with better nutritional status. Female children had a higher risk of death than the males. Mother's education and economic condition of household also showed negative relationships with the risk of death, but the effect of mother's education was modified by economic condition and sex of the children.