The Goldilocks Problem: Comparing a Tight Vs. Adaptive Implementation of the Structured Decision-Making Model

Background and Purpose

The Structured Decision-Making Model (SDM) is an intervention intended to address well-known issues related to subjectivity, bias, and inconsistency in child welfare decision-making through standardization. While the SDM is in use in over 30 states, little is known about whether or not states have implemented this intervention differently and, if so, what the consequences are of distinct implementation strategies. This research compares a tight and an adaptive implementation of the SDM in order to determine the benefits and drawbacks of each approach.

Methods:

A revised institutional ethnography method was used to identify how the SDM was implemented in States A and B. Data was collected through document and policy reviews, observations, and interviews. 66 front line workers (35 in State A and 31 in State B) were interviewed to examine worker’s interpretation of policy and their actual use of the SDM. The data was analyzed using a modified grounded theory approach with special emphasis on the constant comparison method. Using an open coding strategy and aided by NVivo 10, interviews were first analyzed without a particular schema to see what themes emerged. Concepts and categories reflecting recurring themes in the data were identified and a codebook was developed. The author and two Research Assistants then analyzed the data deductively using the codebook.

Findings: 

In State A, the score on the risk assessment determined the trajectory for substantiated cases and workers were mandated to use the SDM as the basis for case decision-making. This tight approach to implementation was largely effective for standardizing decision-making procedures. However, the majority of workers in State A (approximately 2/3) reported broad dissatisfaction with such a strict policy, with 1/3, ignoring it entirely.  

In State B, workers were required to complete the risk assessment but the policy did not make explicit how the results should factor into worker’s case judgments. This adaptive framework allowed for a flexible application of the SDM that explicitly sought to integrate the SDM into local culture and leave room for individual judgment in decision-making. However, a majority of the 31 respondents in State B reported that the SDM had little to no impact on case decision-making.

Conclusion and Implications:

Both tight and adaptive approaches to implementation lead to unique issues.  While State A’s approach is generally effective, the widespread dissatisfaction among workers with their lack of input in relationship to the SDM and, subsequently, case decisions, created problems related to worker burnout and retention.  While State B’s workers were satisfied with their work process, the flexibility allowed them by the implementation strategy meant the SDM functionally had no impact on case decisions. There is a need to develop better models for implementation of the SDM that are able to incorporate flexibility, without becoming so adaptive that they become meaningless, and ones that incorporate fidelity, without alienating the workers who must use them.