Evidence for “Tailoring” in the Matching of Integrated Services to Students’ Developmental Needs in City Connects Schools Using Pattern Analysis and Latent Class Analysis

Abstract

With an increase in emphasis on individual uniqueness and multi-contextual influences, developmental and intervention/prevention science along with similar fields of research (e.g., personalized medicine, personalized learning, health communication, business marketing) have promoted the design and implementation of interventions that would tailor responses and strategies to optimize targeted outcomes based on individual needs and variability (Joyner & Paneth, 2019; Kreuter et al., 1999; Vesanen, 2007). However, in spite of the effort and resources invested in personalization in the past decades, evidence for the realization and utility of tailored interventions have been more anecdotal than quantitatively empirical. The majority of person-centered studies have been qualitative (Lerner et al., 2019). While there is little agreement on what “tailoring” means across the different fields of study, there is a consensus that the term “tailoring” and tailoring-related terms (e.g., personalization, individualization, differentiation, and customization) lack a common and feasible theoretical foundation. Consequently, this semantic crisis has made the construct increasingly difficult to conceptualize and operationalize (e.g. Economist Group, 2021; Shemshack & Spector, 2020). Drawing on insights from the Specificity Principle, Orthogenetic Principle, and Developmental Contextualism in developmental science, this dissertation proposed a provisional definition of “tailoring”: the process of matching unique patterns of services based on each student’s cumulative strengths and needs and the availability of services (e.g., Bornstein, 2015; Lerner et al., 1998; Walsh et al., 2002; Werner & Kaplan, 1956). Guided by this definition, this dissertation sought to find evidence of “tailoring” in one “whole-child,” school-based/evidence-based Integrated Student Support (ISS): City Connects. City Connects partners with school personnel and multiple community agencies to systematically and cost-effectively allocate services/resources to students and their families from low-income communities in order to promote strengths, address needs, and mitigate risks (Moore & Emig, 2014; Dearing et al. 2016; Walsh & Theodorakakis, 2017). After establishing a theoretically-informed basis for “tailoring” as an operationalizable construct, this dissertation employed a comprehensive, three-dimensional approach to data analysis: nomothetic (for finding general/ “universal” trends), differential (for finding differences between groups), and idiographic (for finding differences between individuals) (e.g., Lerner et al., 2019; Overton, 2015; Salvatore & Valsiner, 2010). This was to magnify the descriptive power of the data and findings. In order to accomplish this, the two exploratory substudies in this dissertation employed 1) descriptive analysis, 2) a novel approach for comparing the service patterns matched to each student’s unique sets of strengths and needs, and 3) Latent Class Analysis (LCA). The major findings suggest that “tailoring” in City Connects schools is occurring in five ways: 1) students with higher needs receive more support than students with fewer needs; 2) City Connects is adaptive in responding to the emerging needs of individuals as circumstances change in the course of time; 3) there are unique patterns of services that are either shared (two more students have the same combination of services/types of services) or unshared (only one student has a particular service pattern); 4) service patterns are related to students’ developmental needs (i.e., higher risk level->higher percentages of individualized service patterns); and 5) service pattern matching is purposeful and does not occur randomly. The implications that these findings have on theory, research, and practice are discussed.