• Summary
• Downloads
• Explanatory Notes
• Related Information
• Past Releases

FACTORS INFLUENCING EARLY CHILDHOOD DEVELOPMENT IN TASMANIA - RESULTS OF REGRESSION ANALYSIS

This appendix presents the results of logistic regression analysis to assist with better understanding the extent to which particular personal, family and household characteristics impact on a child's developmental vulnerability in their first year of school. The regression models the influence of selected variables from the ABS Census of Population and Housing and the National Early Childhood Education and Care Collection on Australian Early Developmental Census (AEDC) measures. The results are further discussed in Factors influencing early childhood development in Tasmania article.


LOGISTIC REGRESSION ANALYSIS

Logistic regression is a popular and widely used statistical technique for modelling categorical outcomes; in this case the probability of whether a child is classified as developmentally vulnerable on any of the AEDC domains.

Logistic regression models are generally expressed in terms of the odds of an event occurring when a factor or set of factors are at play. For each factor a comparison group is selected. In the context of this paper, the "odds ratio" for each category represents the likelihood of a child in that group being classified as developmentally vulnerable, in contrast to a child in the comparison group. In the tables below, the significance of a result is indicated by one or more asterisk(*). Where a result does not include an asterisk, the odds ratio is not significant and should be interpreted with caution.

Values below one indicate a decreased likelihood of being classified as developmentally vulnerable on the relevant domain, while values above one show an increased likelihood of being developmentally vulnerable on that domain, relative to the comparison group. Putting multiple variables into a logistic regression model allows analysis of the effect of a factor while controlling for the effects of all other variables in the model, thereby accounting for interactive effects.

The regression model was run separately for each of the five AEDC domains. There were no differences to the model used for each domain.

It is important to note that the effect of the selected factors on early childhood development may be mediated by a range of other contextual, family and individual characteristics. Some of the other factors which are not explored in this regression analysis may include factors such as a child's inherent ability, family cohesion, teacher quality and community support.


RESULTS

The extent to which particular characteristics influence the likelihood of a child being classified as developmentally vulnerable for the Physical health and wellbeing, Social competence, Emotional Maturity, Language and cognition and Communication and general knowledge domains are presented in Tables 1 and 2. The information below discusses the results for selected variables which were statistically significant.

After controlling for the effects of other variables included in the model, the characteristic with the most consistent impact on a child's likelihood of being classified as developmentally vulnerable across all domains was the regularity with which a child was read to, or encouraged in their reading, at home. Across the domains, children who were not read to regularly were 2.8 to 3.9 times more likely to be developmentally vulnerable than children who were read to somewhat regularly. Children who were read to very regularly were 74% to 85% less likely to be developmentally vulnerable than children who were read to somewhat regularly.

Similarly, whether a child's parent(s) or caregiver(s) were actively engaged with the school in supporting their child's learning had a strong and consistent relationship with the child's developmental vulnerability. This item was highly correlated (70%) with whether the child was read to, so could not be included in the same regression model. When running a separate regression model substituting parental engagement for the regularity with which a child was read to, children whose parents were not engaged with the school were 2.1 to 3.2 times more likely to be developmentally vulnerable than children whose parents were somewhat engaged. Children with parents that were very engaged with the school were 75% to 85% less likely to be developmentally vulnerable than children with parents that were somewhat engaged.

A child's sex was also significantly related to their developmental vulnerability, with males being largely more likely to be developmentally vulnerable than females across all five domains. This effect was largest in the Emotional maturity domain, where males were nearly five times (4.7) more likely to be vulnerable than females.

Earlier regression models also included tenure type, need for extra bedrooms, whether the child was cared for by a relative in the year before starting school, and type of internet access. None of these factors were found to have a significant relationship with developmental vulnerability on the AEDC and were subsequently removed.

Characteristics		Physical health and wellbeing	Social competence	Emotional maturity	Language and cognition	Communication and general knowledge
Hours enrolled in preschool per week
	1 to 10 hours	1.56	2.08**	1.89*	1.51	1.64
	11 to 14 hours (comparison group)	-	-	-	-	-
	15 to 19 hours	1.49**	1.60**	1.33	1.46*	1.08
	20 hours or more	2.59**	1.22	0.81	<0.001***	0.72
Whether child was regularly read to, or encouraged in their reading, at home
	Not read to regularly	3.87***	2.81***	2.90***	3.33***	3.09***
	Read to somewhat regularly (comparison group)	-	-	-	-	-
	Read to very regularly	0.21***	0.16***	0.26***	0.15***	0.25***
Sex
	Females (comparison group)	-	-	-	-	-
	Males	1.81***	2.31***	4.68***	1.72***	2.21***
Indigenous Status
	Non-Indigenous (comparison group)	-	-	-	-	-
	Aboriginal and Torres Strait Islander	0.98	0.83	1.25	1.48*	0.99
Child type
	Natural or adopted child	0.87	0.87	0.51**	1.35	2.40*
	Other child type (a) (comparison group)	-	-	-	-	-
Age started school
	Younger than 5.5 years	1.47**	0.92	1.07	0.51***	1.27
	5.5 – less than 6 years (comparison group)	-	-	-	-	-
	6 years or older	0.78	0.76	0.89	1.01	0.88
Highest parental education
	Postgraduate or Graduate	1.06	0.50	0.43*	0.34*	0.28**
	Bachelor Degree	0.87	0.83	0.55*	0.60	0.63
	Advanced Diploma & Diploma	0.76	0.68	0.79	0.67	0.47*
	Certificate III & IV	0.94	0.89	0.95	0.83	0.67
	Year 12	1.16	1.12	0.96	1.25	1.25
	Year 11 or below (b) (comparison group)	-	-	-	-	-
Parental labour force status
	All employed (comparison group)	-	-	-	-	-
	Partial employed (c)	1.27	1.36	1.10	1.71*	1.63*
	All unemployed	0.94	1.15	0.83	1.36	0.98
	All not in labour force	1.27	1.34	1.10	1.53	1.65*
Number of children in family
	One	1.14	1.33	1.25	0.98	1.04
	Two (comparison group)	-	-	-	-	-
	Three	0.74*	0.64**	0.47***	0.89	0.79
	Four	0.61*	0.61*	0.45**	0.64	0.59*
	Five or more	0.87	0.46**	0.55*	0.48**	0.46*
Birthplace of parents
	Both parents born in Australia (comparison group)	-	-	-	-	-
	Both parents born overseas	0.35*	1.03	1.47	0.70	1.47
	Father only born overseas	1.38	0.61	0.63	1.26	1.11
	Mother only born overseas	0.80	1.37	1.73	0.90	1.06
Family type
	Married (d) (comparison group)	-	-	-	-	-
	Defacto (d)	0.86	0.89	0.81	0.91	1.17
	Lone parent	0.98	1.07	1.00	1.20	0.77
Mother's birth age (e)
	15 to 19	0.85	1.12	1.22	0.49	0.71
	20 to 24	0.80	0.98	1.08	1.20	1.04
	25 to 29	0.79	0.86	0.83	1.00	0.94
	30 to 34 (comparison group)	-	-	-	-	-
	Over 35	0.94	0.77	1.07	1.35	1.27
Weekly Household Income
	$1 - $599	1.25	1.17	1.32	1.16	1.08
	$600 - $999	0.94	0.80	1.21	1.11	0.73
	$1,000 - $1,499 (comparison group)	-	-	-	-	-
	$1,500 - $1,999	1.10	1.30	1.57*	1.55	0.99
	$2,000 - $2,499	1.31	1.50	1.93*	1.07	0.86
	Over $2,500	1.05	1.52	1.65	1.40	0.88
SEIFA
	Quintile 1 - Most disadvantaged (comparison group)	-	-	-	-	-
	Quintile 2	0.62**	0.65*	1.05	0.95	0.56**
	Quintile 3	0.55**	0.94	1.04	0.80	0.62*
	Quintile 4	0.88	1.04	0.99	0.97	1.19
	Quintile 5 - Most advantaged	0.64	0.92	1.00	0.75	0.90
Remoteness
	Inner Regional (comparison group)	-	-	-	-	-
	Outer Regional	0.76*	0.66**	0.66**	0.84	0.98
	Remote or Very Remote	1.16	1.59	0.48	1.18	1.14
*** Significantly different effect to the comparison group (p<0.001)
** Significantly different effect to the comparison group (p<0.01)
* Significantly different effect to the comparison group (p<0.05)
(a) Includes step-children of either parent, foster children, grand-children, otherwise related children and unrelated children.
(b) Includes Certificate I & II.
(c) Couple families where only one parent was employed, either full-time or part-time.
(d) Opposite sex couples only.
(e) Includes natural and adoptive mothers.
Source: Integrated Tasmanian Education and ABS Census Dataset.

Characteristic		Physical health and wellbeing	Social competence	Emotional maturity	Language and cognition	Communication and general knowledge
Whether parents were engaged with the school
	Parents were not engaged	3.23***	2.05***	2.14***	2.99***	2.76***
	Parents were somewhat engaged (comparison group)	-	-	-	-	-
	Parents were very engaged	0.21***	0.15***	0.24***	0.15***	0.25***