Engaging hard-to-reach families and children
8. Effects of Communities for Children on hard-to-reach families and children
8.1 Defining hard-to-reach families
8.2 Findings
In addition to the qualitative research reported above, analysis of the Stronger Families in Australia (SFIA) dataset was undertaken, to provide insight into the effectiveness of the model for hard-to-reach groups in Communities for Children areas. A detailed analysis of hard-to-reach families in Communities for Children and control communities, as well as subgroups (economically deprived families and those with low parental education) in SFIA is available in the main SFIA report (Edwards et al. 2009). Here we provide a summary of those findings.
Quantitative analysis requires a fixed, working definition of hard-to-reach families. In some ways this classification is at odds with the findings emerging from the literature review, and reinforced in the qualitative findings, that the definition of hard-to-reach is context dependent.
However, there are some characteristics that emerged from the literature and in our interviews with service providers that aided a reasonably accurate operational definition. This is necessarily based on family characteristics only, so does not take into account that barriers to service utilisation and engagement can also arise from the ways services are delivered.
In the quantitative analysis, we compared families considered hard-to-reach using a fixed definition, with others in Communities for Children communities. For the purposes of this analysis we focused on four groups of families:
- the total SFIA sample: As described in Edwards et al. (2009), these consisted of a random sample of approximately 200 children in each of 10 Communities for Children sites and five contrast sites, representing about 40 per cent of children under 5 years in those communities.
- hard-to-reach families, as defined below.
- not-hard-to-reach families: This is also an important group to study because the literature indicates that programs such as Communities for Children often access those families who are easiest to reach in the first instance, and only address the needs of the harder-to-reach families in the longer-term.
- particular groups of hard-to-reach families, in particular families with relatively low incomes and families where parents have low levels of education.
[ top ]
8.1 Defining hard-to-reach families
The hard-to-reach populations were defined according to the following family characteristics, designed to capture the most disadvantaged groups within Communities for Children sites:
- no father: no father present in the household
- jobless household: mother is unemployed or not in the labour force and father not working or not present
- poor parents: parental income is less than $500 per week
- low education: maternal education is Year 10 or less
- Indigenous
- parent born overseas.
Persons in the SFIA sample who met any of these criteria were coded as hard-to-reach. All remaining persons who met none of these criteria were coded as not-hard-to-reach.
Table 2 below gives details of the numbers of families in each hard-to-reach category.
It can be seen from this that the hard-to-reach sample is made up largely of people born overseas, in jobless households, and with no father present (sole parents).
In addition, almost half the people included in the hard-to-reach sample (47.9 per cent) were counted as hard-to-reach on two or more categories. This indicates that these factors were clustered together and that many families were disadvantaged in several ways. Overall the table shows that a significant proportion of families in Communities for Children sites could be defined as hard-to-reach.
Teenage parents were also considered, but ultimately excluded from the hard-to-reach sample. Teenage parents were excluded because of the very small sample size—less than 15 people in the whole balanced panel 6 sample.
Other groups that have been identified as hard-to-reach, including parents with a mental illness, parents who misuse substances and families living with domestic violence, were not included in these definitions because this information is not available in the SFIA dataset.
| No. (at Wave 1) |
% of full Wave 1 sample (balanced panel) |
% of Wave 1 hard-to-reach sample |
|
|---|---|---|---|
| No father present in the household | 396 | 22.0 | 35.6 |
| Jobless household | 387 | 21.5 | 34. |
| Poor parents (<$500) | 295 | 16.4 | 26.5 |
| Low education (educational attainment of Year 10 or less) | 315 | 17.5 | 28.3 |
| Parent born overseas | 559 | 31.0 | 50.2 |
| Indigenous | 133 | 27.4 | 44.5 |
| In at least two categories | 533 | 29.5 | 47.9 |
[ top ]
8.2 Findings
Most of the significant findings of the analysis of hard-to-reach and not-hard-to-reach groups in SFIA are in line with the analysis of the whole SFIA sample.
The main findings according to the difference-in-difference estimation are:
- The number of jobless households declined significantly in hard-to-reach households in Communities for Children sites compared to control sites, matching findings from the full sample and differential analyses.
- Several outcomes were significantly different in hard-to-reach households located in Communities for Children sites compared to control sites, including worse child and parental physical health, and improved involvement in community service activity.
The main findings of the cross-sectional analysis of the hard-to-reach subsample at Wave 3, which compared Communities for Children and control sites, were that:
- Child physical health was significantly worse in Communities for Children sites compared to control sites, matching findings from full sample and differential analyses. This is a likely effect of reporting improvements.
- Hostile parenting had improved—it was significantly lower in Communities for Children sites compared to control sites, matching findings from full sample analyses.
- Jobless households had improved to a degree that was nearly significant in Communities for Children sites compared to control sites.
The main findings of the cross-sectional analysis of the not-hard-to-reach subsample at Wave 3, which compared Communities for Children and control sites, were that:
- Hostile parenting had improved—it was significantly lower in Communities for Children sites compared to control sites, matching findings from full sample analyses.
- Parental self-efficacy was significantly higher in not-hard-to-reach households in Communities for Children sites compared to control sites.
- Unmet service needs were worse—they were significantly higher in Communities for Children sites compared to control sites, a finding related to the exclusion from the not-hard-to-reach sample of parents born overseas. This suggests that native English speakers are more aware of, and better able to articulate, service gaps.
- Wave 3 showed improvements in child emotional and behavioural problems, along with minor but significant improvements in receptive vocabulary and verbal ability for children in hard-to-reach families in Communities for Children sites as opposed to contrast sites.
Discussion
Overall this analysis confirms the main conclusions in the National evaluation (2004–2008) of the Stronger Families and Communities Strategy 2004–2009 report (Muir et al. 2009)—that the effects of the Communities for Children initiative were generally in the positive direction, with a few outcomes beginning to show significance and a very few negative findings (for example, where Communities for Children sites appear to be doing worse than contrast sites). Some of these latter findings can best be explained by increased parental sensitivity and awareness of their own and their children’s health status and their access to services.
The findings are important as they show that the Communities for Children initiative was as effective in addressing the needs of hard-to-reach families as those who were not deemed hard-to-reach, confirming that the active steps which Communities for Children sites took to engage with these families were paying off in terms of outcomes for children. Particularly encouraging was the finding that community participation increased for the hard-to-reach group, as this was a high priority for Communities for Children and was seen as a very challenging objective.
As with the main analysis, however, it is important to consider the major trends for families who may be hard-to-reach rather than focusing on specific outcome areas. This is partly because these are very early outcomes, and so the fact that some findings are statistically significant while others are less than significant could be a factor of the timing of the data collection rather than of the impact of Communities for Children on that specific outcome. Had Wave 3 data been collected six months later, it is very likely that more of the outcomes would have been significant. This is because families and children in the community would have been exposed to Communities for Children for a longer period.
These findings are also consistent with the qualitative findings in relation to the contextual nature of being hardto- reach. They show that some outcomes were only significant for some sections of the hard-to-reach population and not others. Bearing in mind the discussion above—that it is premature to pay too much attention to specific findings—it may yet be the case that some aspects of the intervention had different effects for some subgroups (for example, families in which parents were born overseas). It may also be the case that specific definitions of hard-to-reach will affect outcomes, so that the effect of the intervention on hard-to-reach families, to some extent, depends on how hard-to-reach is defined and used in the analysis. In general, outcomes for the whole hard-to-reach group were similar to those of the specific subgroups that were studied in most depth —low-income families and families where the mother had low levels of education. Furthermore, in broad terms the outcomes for hard-to-reach families were similar to those of the total population of families in the Communities for Children communities. This indicates that Communities for Children was equally effective for hard-to-reach and not-hard-to-reach families, although the effects of the initiative were slightly different for the two groups.
It also shows that, although the specific definition of hard-to-reach may affect particular outcomes, it does not really change the broader picture of the effectiveness of Communities for Children for hard-to-reach families.
[ top ]