Testing for seasonality in Canadian child welfare investigations. (March 2021)

• Record Type: Journal Article
• Authors: Shields, Margot; Tonmyr, Lil; Morin, Yves; Hovdestad, Wendy; Adams, Nicole; Esposito, Tonino; Jervis, Laurie; MacIsaac, Scott; Pang, Cheuk; Peterson, Michael; Thompson, Diane; Chabot, Martin; Trocmé, Nico
• Is Part Of: Children and youth services review. Volume 122(2021)
• Publication Date: 2021-03
• Subjects: Bias -- Child abuse and neglect -- Child emotional maltreatment -- Child neglect -- Child physical abuse -- Child protection -- Seasonal patterns -- Child sexual abuse -- M7 statistic; Social work with children -- Periodicals; Social work with youth -- Periodicals; Adolescent -- Periodicals; Child Welfare -- Periodicals; Social Work -- Periodicals; Service social aux enfants -- Périodiques; Service social à la jeunesse -- Périodiques; Electronic journals; 362.705
• DOI: 10.1016/j.childyouth.2020.105878
• Languages: English
• ISSNs: 0190-7409

Abstract

Highlights: Statistically significant evidence of seasonality was observed based on administrative data from child welfare agencies from six Canadian provinces/territories for the years 2005–2015; percentages of child maltreatment (CM) investigations occurring in Quarter (Q)3 (July-Sept.) were relatively low. Q4 (Oct.-Dec.) provides a good representation of expected volume of CM investigations across the year; however, sexual abuse and neglect investigations were less likely to occur in Q4 and physical abuse and emotional maltreatment investigations were more likely to occur in Q4. Annual estimates based on administrative data from Q4 using CIS estimation procedures resulted in minimal biases for investigated and substantiated CM. However, due to the unique patterns observed for Q4 for specific types of CM, there was a tendency for sexual abuse and neglect investigations to be underestimated (particularly sexual abuse) and for physical abuse and emotional maltreatment to be overestimated (particularly physical abuse). In many cases the observed biases for specific types of CM exceeded 5%. Abstract: The Canadian Incidence Study of Reported Child Abuse and Neglect (CIS) derives annual estimates of child maltreatment (CM) based on cases of CM reported to child welfare agencies in the fourth quarter (October, November, December) of the year the CIS is conducted. The present study investigates seasonal patterns of CM investigations and examines potential biases in CIS estimates that result from annual estimates based on quarter 4 (Q4) data. Seasonal patterns by quarter are examined based on child welfare administrative records, from six Canadian provinces/territories for 11 years (2005–2015). CIS estimation procedures are applied to this administrative data to assess potential biases in deriving annual estimates based on Q4. Based on the administrative data, there was evidence of a statistically significant seasonal effect by quarter with relatively fewer cases being observed in Q3 (July, August, September). Using CIS estimation procedures, annual estimates derived from Q4 administrative data indicated a tendency for sexual abuse and neglect investigations to be underestimated, and physical abuse and emotional maltreatment investigations to be overestimated. The CIS has been and will continue to be used by researchers and policy makers to establish the need for and evaluate the success of programs designed to prevent CM in Canada. However, methodological improvements would further enhance the capability of the CIS to derive accurate annual estimates of specific types of CM among Canadian children.