Research report no.4 2000
Pathways from infancy to adolescence: Australian Temperament Project 1983-2000
by Margot Prior, Ann Sanson, Diana Smart, and Frank Oberklaid
2. Beginnings
How did we enrol the ATP sample?
The families enrolled in the study came from 67 local government authorities across the whole state of Victoria, selected upon the advice of the Australian Bureau of Statistics, to give a true and unbiased representative sample of the whole population of Victoria. Twenty of the selected areas were urban (1604 children) and 47 were rural (839 children). This paralleled the urban/rural population balance in the state in 1983 when we began the project.
Method of recruitment of the study sample
We were able to contact the families with the assistance of the Maternal and Child Health Division of the Victorian Health Commission in 1982. The Infant Welfare Centres (IWCs) which are provided by this service (now called Maternal and Child Health Centres) establish contact with 94 per cent of families with a new infant in Victoria.
The mother of every 4-8-month-old infant who attended an IWC in one of the selected local government authority areas in the two weeks between 22 April and 6 May 1983 was handed an Australian Temperament Project questionnaire by the Infant Welfare Sister for completion and return in a pre-paid envelope.
The Infant Welfare Sisters also completed a brief questionnaire covering the child's birth history, current weight, type of feeding (breast or bottle), a rating of each child's temperament (from much easier than average, through average, to much more difficult than average), and a rating of their perception of the current adjustment of the mother-baby pair. They also let us know if they thought that any of the mothers had difficulties completing the questionnaires because of reading or English language problems.
Some sample characteristics
The total number of children involved at the first (infancy) stage of the research was 2443. Fifty-two per cent were male and 48 per cent were female.
Table 1 summarises the characteristics of the original Australian Temperament Project sample.
- There were 28 sets of twins.
- Average birth weight of the infants was 3400 grams.
- One hundred and twenty-six infants were pre-term, that is, born at 36 weeks gestation or less (5.7 per cent of sample); of these, only one had birth-weight less than 1000 grams, while five were below 1200 grams.
- Forty-three per cent were being breast fed, 30 per cent were bottle fed, and 23 per cent had begun with breast feeding but had moved to bottle. About 5 per cent of mothers were using both methods of feeding at the time of the infant survey. Mothers provided ratings of three key infant difficulties.
These indicated that:
- almost 30 per cent of infants had moderate or severe colic problems;
- 17 per cent had moderate or severe sleep problems; and
- 13 per cent showed moderate or severe difficulties with excessive crying.
Survey points across age
From infancy onwards, families have been contacted with requests to complete questionnaires at roughly 15-18 month intervals. Survey points and numbers successfully followed-up at each stage are listed below.
- 1983: In infancy, 2443 families were initially contacted and provided the sample cohort for our longitudinal study.
- 1984: In early toddlerhood, when children were 18-24 months, two-thirds of the sample were asked to complete questionnaires and 1280 families (of a possible 1453 families) did so.
- 1985: The remaining one-third, plus one-half of the 1984 sample were asked to complete questionnaires when the children were 32-36 months of age and 1360 did so.
- 1986: When the children were 3-4 years old (pre-school age); 1716 families.
- 1988: When the children were 5-6 years old (usually Preparatory Grade); 1721 families.
- 1990: When the children were 7-8 years old (usually Grade 2); 1605 families.
- 1992: When the children were 9-10 years old (usually Grade 4); 1545 families.
- 1994: When the children were 11-12 years old (usually Grade 6); 1470 families.
- 1995: When the children were 12-13 years old (usually Year 7); 1280 families.
- 1996: When the children were 13-14 years old (usually Year 8); 1400 families.
- 1998: When the children were 15-16 years old (usually Year 10); 1375 families.
- 2000: When the children were 17-18 years old (usually Year 12); data currently being collected.
Because of the very large sample size and limited research budgets, all of these wholesample surveys have been by mail. Parents have received a package of questionnaires and rating scales which we have asked them to complete about their project child and return to us. It is usual in studies of this kind for families to have missed at least one survey point due to family pressures at a particular time. Nevertheless, our response rate has been around 80 per cent at every survey wave, which is a very good participation rate.
Neither we nor the families enrolled in the project in 1983 realised at the time we began that we would continue the study for so many years. But since we had developed such an excellent representative sample and were able to learn more and more about the children, we realised how important it was to continue on through the school years, and to see how the children developed in later childhood and adolescence.
Approximately 15-20 families moved each month, at least in the early years of the project. By the time the children were 9-10 years of age, two-thirds of families had moved house at least once; 17 per cent had moved three or four times and 3 per cent had moved five or more times. We have managed to maintain contact with most of these families, but some became 'lost' to us as they moved house. There have also been some families who have decided to withdraw from the study during the 18 years of its existence. In this group there has been a somewhat greater loss of families with non-Australian parents, and less-advantaged circumstances. However, two-thirds of the original sample are still participating and they are broadly similar in family characteristics to the original complete sample. Nevertheless, we are aware that some of the families now lost to the study are those who have experienced difficulties of one kind or another. This situation is consistent with the experience of similar studies across the world. Our retention of 67 per cent (approximately 1650 families) is an excellent rate of commitment to the research programme, which we know asks a great deal of the project families.
In the first sampling stage, approximately 45 per cent of families were non-metropolitan residents. In the 1994 and following surveys, we categorised our sample into four groups, as a way of capturing the differing areas of residence. These were: Melbourne metropolitan, outer metropolitan (for example, Belgrave, Melton, Cranbourne), large provincial cities (such as Bendigo, Ballarat), and rural.
In 1994, when children were aged 11-12 years, 53 per cent lived in the metropolitan area, 8 per cent in outer metropolitan areas, 8 per cent in large provincial cities, and 31 per cent in rural areas. This corresponds well to the population of Victoria as a whole. A number of families in our sample are now living interstate or overseas, but remain enrolled in the study and continue to complete questionnaires. In 1994, 90 per cent of the sample were living in owner-occupied homes, a very high proportion of the community by comparison with the situation in most other countries.
Some figures on unemployment and family constellation in the 1990s
Unemployment: In the following years, the following percentages were reported as unemployed and looking for work.
- 1992 (children aged 9-10 years); 11 per cent of fathers and 9 per cent of mothers;
- 1994 (11-12 years) 8 per cent fathers, 7.5 per cent mothers;
- 1995 (12-13 years) 9 per cent fathers, 7.5 per cent mothers;
- 1996 (13-14 years), 9 per cent fathers, 8 per cent mothers;
- 1998 (15-16 years) 7 per cent fathers, 6 per cent mothers.
Separation and divorce: At each survey point, we asked parents to report on their marital and family status. By 1994 (late childhood), approximately 15.6 per cent of children were not with both biological parents. Eighty per cent of these children were with their biological mothers, either in a single parent family or with mother and a stepfather. The number of families where it was reported that parents had separated or divorced went from 47 in 1986, gradually increasing at each survey point to 152 families by 1998. A current study is focused on the reactions and adjustment of project children whose parents have separated or divorced and this will be reported in the near future.
Teacher contributions
From the pre-school stage onwards, we asked parents to let us know the name of the child's teacher, and to give us permission to contact the school so that we could gather information about the child's adjustment and progress in school. We then sent packages of questionnaires relevant to the age and stage of enrolled ATP children to the teachers in schools across the state (and sometimes interstate and overseas). This has given a more complete picture of children's development.
At the secondary school level, we ceased sending questionnaires to teachers because the secondary system involves each child having different teachers for different subjects. It is harder to identify a teacher who might know an individual child sufficiently well to be able to report on temperament, behaviour, peer relationships, and school progress.
Child self-reports
At each survey point from the age of 11-12 years (usually Grade 6) we have asked the children to complete questionnaires, so that we can discover their views of themselves at this point in their lives. Some of their responses have been delightful; besides their questionnaire responses some children have sent photographs, poems, letters, and short biographies, and in a later section we give some examples of how our project children have seen themselves. Since the children reached secondary school, we have routinely included them as respondents in all surveys, as informants on themselves.