Using Predictive Evaluation to Design, Evaluate, and Improve Training for Polio Volunteers

Abstract

BACKGROUND: Predictive Evaluation (PE) uses a four-step process to predict results then designs and evaluates a training intervention accordingly. In 2012, the Sustainable Management Development Program (SMDP) at the Centers for Disease Control and Prevention used PE to train Stop Transmission of Polio (STOP) program volunteers. METHODS: Stakeholders defined specific beliefs and practices that volunteers should demonstrate. These predictions and adult learning practices were used to design a curriculum to train four cohorts. At the end of each workshop, volunteers completed a beliefs survey and wrote goals for intended actions. The goals were analyzed for acceptability based on four PE criteria. The percentage of acceptable goals and the beliefs survey results were used to define the quality of the workshop. A postassignment adoption evaluation was conducted for two cohorts, using an online survey and telephone or in-person structured interviews. The results were compared with the end of workshop findings. RESULTS: The percentage of acceptable goals across the four cohorts ranged from 49% to 85%. In the adoption evaluation of two cohorts, 88% and 94% of respondents reported achieving or making significant progress toward their goal. A comparison of beliefs survey responses across the four cohorts indicated consistencies in beliefs that aligned with stakeholders' predictions. CONCLUSIONS: Goal statements that participants write at the end of a workshop provide data to evaluate training quality. Beliefs surveys surface attitudes that could help or hinder workplace performance. The PE approach provides an innovative framework for health worker training and evaluation that emphasizes performance.

Cite this work

No public copy is stored for this record, so only the metadata is cited.

Stored only when both papers are in the corpus — 1 citations exist upstream.

Papers that cite the same references — the relational fallback while embeddings are still backfilling.