In their April 8 talk, Professor Michal Perlman and her team highlighted the challenges in understanding parents’ perspectives of early childhood education and care services. Traditional survey methods are not well-suited for the task. “When we ask parents questions about their child care provider, the results tend to be very positively skewed,” Perlman explained. “Many more parents report that their children are in licensed home care than is possible. It is really difficult to interpret results from parent surveys.”
Undeterred, Perlman searched for new research tools. “We had the good fortune to meet Delaine Hampton from the Rotman School Management, who introduced us to methods used in marketing research known as conjoint survey methods,” Perlman said. Surveys that consider different options together or “jointly” allow for more nuanced insights about what people choose when they have an array of options and choice is constrained in various ways. Hampton explained that “conjoint survey methods reveal preferences and show how people trade off one feature for another when they can’t have everything.”
Collectively, team members Adrienne Davidson, Samantha Burns, and Sumayya Saleem, demonstrated how conjoint survey methods are particularly suited to studying parental choices around early childhood education and care.
In surveying over 700 parents in partnership with the City of Toronto and EarlyON Centres through conjoint methods, Davidson pinpointed attributes parents valued, including cost, location, level of caregiver training, physical space, flexibility in hours, and caregiver interaction. Next, she used a data analysis technique called latent profile analysis to identify groups of parents who had similar profiles of preferences. She then mapped parents’ preferences onto aspects of their background and life situation. Combined, these methods allowed Davidson and the team to predict whether parents in a particular demographic will lean towards clusters of choices.
While Davidson focused on parents’ preexisting views, Burns asked what happens when parents are provided with additional information about child care settings. To what degree does additional information impact parents’ choices? Using conjoint methods combined with an experimental design, Burns found that respondents with lower incomes and less formal education had an even stronger preference for licensed home care when given a description of what licensing entails.
Saleem brought the discussion full circle by reconsidering the utility of asking parents qualitative/open ended questions when used alongside conjoint survey methods. What can we know about the differences between preferences stated in response to open-ended survey questions and preferences revealed through conjoint survey methods? Saleem’s work indicates that stated and revealed preferences were more closely correlated for parents with higher incomes.
Perlman traced a common thread throughout their papers: the importance of addressing equity issues. “Parents whose children could benefit most from good quality care are in the weakest position to secure it for their children,” she said. Another significant finding is an overwhelming preference for licensed care, especially when parents are provided information about licensing. “We need to attend to this finding from a policy standpoint,” Perlman said. “The quality of the care children receive matters, especially for children from lower social economic status backgrounds, and parents are the ones making the key decisions about early childhood education settings.”
For additional information: Davidson, A., Burns, S., White, L., Hampton, D., & Perlman, M. (2020). Child care policy and child care burden: Policy feedback effects and distributive implications of regulatory decisions. Journal of Behavioral Public Administration, 3(2).
This talk focused on emerging research; additional publications are forthcoming.