Early Development Instrument
The Early Development Instrument (EDI) is a population-based measure of children's ability to meet age-appropriate developmental expectations developed by the Offord Centre for Child Studies at McMaster University. The EDI is useful for communities, organizations and policy-makers to understand vulnerability and differences in developmental outcomes, and to monitor changes over time. Below you will find more information about Toronto's 2014-2015 results. The Offord Centre also reports on how Toronto children are doing compared to the Ontario population.
One or more domains: 29% Two or more domains: 14%
Vulnerable in one or more domains
In Toronto in 2014/2015, 28.7% of children were vulnerable on at least one domain. This is slightly better than the results for all of Ontario, where 29.4% of children were vulnerable on one or more domains. Figure 2a shows that the percent of children who are vulnerable in one or more domain varies by neighbourhood. Figure 2b shows that there are clusters of neighbourhoods where the percent of children who are vulnerable on one or more domains are significantly higher and lower than Toronto overall.
Vulnerable in two or more domains
In Toronto in 2014/2015, 13.6% of children were vulnerable on two or more domains. This is slightly less than the 14.4% of Ontario children who were vulnerable on two or more domains. Figure 3a shows that the percent of children who were vulnerable on two or more domains varied by neighbourhood across the City. Figure 3b shows the neighbourhoods that had significantly a higher or lower percent of children who are vulnerable on two or more domains, compared to Toronto overall.
Inequalities in Vulnerability
There are important inequalities in vulnerability by socio-demographics.
Figure 4: Percent of children who are vulnerable on two or more domains by gender, Kindergarten students, Toronto, 2014/2015.
In Toronto in 2014/2015, Male children were significantly more likely to be vulnerable in early development. Figure 4 shows that 18% of male children are vulnerable on two or more domains, compared to 9% of female children.
Figure 5: Percent of children who are vulnerable on two or more domains by age groups, Kindergarten students, Toronto, 2014/2015.
Younger children were significantly more likely to be vulnerable, compared to older children. Figure 5 shows that 18% of the youngest children were vulnerable, compared to 14% of children in the middle age group and 9% of the oldest children.
Figure 6: Percent of children who are vulnerable on two or more domains by ELL/FSL status, Kindergarten students, Toronto, 2014/2015.
English Language Learners (ELL) and French as a Second Language (FSL) children were significantly more likely to be vulnerable than their non-ELL/FSL peers. Figure 6 shows that 28% of ELL/FSL children are vulnerable, compared to 12% of non-ELL/FSL children.
Figure 7: Percent of children who are vulnerable on two or more domains by income quintile, Kindergarten students, Toronto, 2014/2015.
There is a socio-economic gradient in vulnerability in early development. Figure 7 shows that children in the lowest income quintile were significantly more likely than children in the highest income quintile to be vulnerable on two or more domains.
How is the EDI measured?
The EDI is a short 103-question survey filled out by Kindergarten teachers in the second half of the school year. Teachers complete the EDI for each child in their class. Based on a combination of questions, children receive a score from 1 to 10 on each domain. Children who meet all age-appropriate developmental expectations of a specific domain would receive a 10. The vulnerability cut-off is based on the lowest 10% of children in the first provincial EDI collection in Ontario from 2004 to 2006. When analyzing the EDI data, children with special needs, those not in class for at least one month, and those with missing scores for at least two domains were excluded.
- Figure 2a and 3a use quartiles to display vulnerability by neighbourhood. Neighbourhoods are characterized into 4 groups of 35 neighbourhoods each.
- In Figure 2b and 3b, significant differences were estimated using overlapping 95% confidence intervals estimated using the Normal distribution. Although this method is conservative and most appropriate when comparing mutually exclusive groups, it was chosen as an objective means of making conclusions on population-based data.
- In Figure 5, children were categorized into 3 groups with one third of the children in each. The “youngest” group represents the youngest one third of children, while the “oldest” group represents the oldest one third of children. The “middle” group contains the remaining third of children.
- In Figure 6, ELL and FSL students were combined due to the relatively small proportion of FSL students in Toronto (1.5%).
- Figure 7 uses an income quintile analysis to estimate inequalities by socioeconomic status. It uses five groups, each containing approximately 20% of the population, were created by ranking Toronto's census tracts based on the percent of residents living below the Statistics Canada after-tax Low Income Measure (LIM). The "Lowest Income" group includes the census tracts with the highest percent of people living below the LIM. The "Highest income" group includes the census tracts with the lowest percent of people living below the LIM. LIM is an income level set at 50% of the median income in Canada in a given year, adjusted for household size.
EDI: Ministry of Education. Early Developmental Instrument. 2014/2015.
Income Quintiles: Statistics Canada. Income Estimates for Census Families and Individuals (T1 Family File), Table F-18. 2013.