ECDC's logo
The thousand-family research study

The most important study undertaken by the Early Childhood Development Centre was an analysis of the outcomes of a two-year home visiting intervention programme in seriously disadvantaged areas in six health authorities - four in England, one in Wales and one in the Republic of Ireland.

A total of 1,031 families comprised the initial intervention and control samples (678 and 373 respectively). Each of the intervention families was visited monthly by one of a team of 86 programme-trained health visitors over a two-year period, while the control families received normal health visiting support. (In the Republic of Ireland it was triple duty public health nurses who provided home visiting support.) The research team of 10 people included field interviewers, nutrition researchers, computer programmers and a leading statistician.

A number of disadvantaged areas in each of the six health authorities were identified by nurse managements in the relevant health authorities. Geographically-based health visitors in each individual area were paired by managers with the aim of ensuring that the housing estates covered approximately similar levels of disadvantage, and were then randomly allocated as intervention or control. Each health visitor, intervention and control, was asked to prepare a list of all the families in her caseload who had children aged between 3 and 27 months. Families were randomly selected from those lists for inclusion in the intervention or control samples, together with their target age child.

The training of the programme visitors included extensive orientation and guidance before the start, focusing primarily on the empowering philosophy of the programme, and on the effective use of the programme materials, including illustrated information ‘cartoons’ and visiting forms. Subsequently each programme visitor received an hour’s individual training, monitoring and support each month for the two years that the programme visiting continued. At the end of the programme the control visitors were offered insights into the nature of programme visiting and its philosophical basis.

A team of four highly trained interviewers undertook all the research interviews and assessments in each of three years - i.e. at the start, middle and end of the programme period for each family. Each assessment visit to the home lasted approximately two hours, with suitable breaks for coffee and informal chats. By way of thanks each family, intervention and control, was given a photograph of their child after each visit. For logistical reasons the interviewing took place on a rolling schedule. It took each interviewer approximately a year to complete all the families in each of the three annual rounds of visiting. More details on the design of the programme and its evaluation are given in an extensive document referenced at the bottom of this summary.

In general the findings were positive. It was concluded that, measured by a variety of statistical and other indicators, the programme as it was carried out in the first phase of the Child Development Programme was a considerable success, bringing about major changes in the home environments and moderate changes in the developmental levels of the target children, when compared with controls. Some of the principal findings are as follows:

Sample     1,051 children (this total includes twenty pairs of jointly targeted twins) were recruited for the study, together with their parents; attrition losses over 32 months totalled 33 per cent; an assessment of the attrition samples showed no differential loss in the development scores of cases across intervention and control samples for three combined areas, but a loss favouring controls in a fourth area.

Validity and reliability    Assessment of the validity and reliability of some of the key variables showed reasonably acceptable limits, with split-half reliabilities ranging from 0.41 to 0.93.

Parental home environment     Scores on most of the parental home environment variables changed in favour of the intervention samples between years 1 and 3, a number of these variables showing significant and highly significant changes when comparing intervention and control.

Fewer intervention families in bottom 25%    Assessments of the number of families below the 25th percentile on the home environment and child development variables showed that in most cases there were smaller percentages of intervention families in the bottom groups in year 3 than there were in year 1. This is an important finding, indicating a reduced level of potential failure in the intervention children when compared with the controls.

New regression model shows more reliable results overall     Tests on the effectiveness of a relatively new regression model used in the analysis of the programme data showed considerably more reliable results when compared with conventional forms of regression.

Prediction analyses confirm direction of programme effect     Test regressions carried out to check on the directionality of the programme model predictions between year 1 and year 3 variables found that the model was satisfactory and the regressions were acceptable, in predicting strongly from years 1 to 3 and only weakly from years 3 to 1.

Many different results confirm effectiveness of intervention     Regression-based predictions of seven major home environment variables in year 3 showed highly significant contributions from the intervention programme, over and above the predictive strength of all the year 1 variables. The intervention was also shown to predict, at a highly significant level, the child's global development score in year 3. These results applied to three of the Areas. The results in a fourth Area, where the socio-economic status of the samples were strongly biased in favour of the controls, showed little predictive effect for the interven¬tion itself (though positive effects were found in this Area for a number of other measures, including nutritional). Overall these findings, rela¬ting a very large number of variables within multivariate predictive models, are strong evidence for the success of the Child Development Programme methods in altering the children's environments and, in consequence, their developmental levels.

Views of participants     The views of parents, health visitors and managers who had been involved in the programme directly or indirectly were widely supportive of the initiative, with only very few dissenters.

Nutrition findings     Nutritional assessments showed that despite the physical growth of the children in these social stress areas over the two-year period of assessment, the absolute quantities of micro-nutrients in their diets declined, increasing the risk of malnutrition and ill health.

Changes in dietary quality     Nutrient intake for the intervention samples (in the two programme regions which were closely monitored by nutrition researchers as well as by the trained research interviewers) increased relative to the controls, and considerably fewer intervention children fell below the 50th and 75th per cent levels of the recommended daily allowances of various nutrients.

Physical changes     Mean height changes at a level of high significance occurred in one of the two regional intervention samples where dietary intakes were measured; in the other region there were more changes upwards across percentile lines and fewer changes downwards, for anthropometric measures on intervention children, compared with controls.


The above findings provide a brief summary of the evidence for the numerous positive outcomes of the main research programme undertaken in the 1980s. In essence, the various research studies and a great deal of contemporary programme visiting have established that the programme model is effective in improving child outcomes in communities suffering disadvantage, where developmental deficits are evident as a result of environmental factors and the often limited socio-educational background of the parents.

The programme model is analysed in more detail on the next page in this website. That page offers an important overview of programme thinking, especially as the various research outcomes and evaluations described on the present page offer strong evidence in favour of the basic empowerment model.

Ref: Barker, W and Anderson, R (1988) The Child Development Programme: an evaluation of process and outcomes (112 pp)  Early Childhood Development Centre.   This report can be ordered from the ECDC at the address on the website’s Contact page.

Return to page on Research Findings

To print article

Hold Ctrl

Click P