The purpose of this study was to investigate the extent to which the efficacy of the Teaching Early Literacy and Language (TELL) curriculum package for low-income preschoolers was impacted by participant characteristics and variables associated with implementation. In a follow-up to a randomized controlled trial (Gray et al., 2024), data was collected from 174 preschool children whose teachers implemented TELL, a Tier 1, whole-class curriculum package targeting the improvement of oral language and early literacy skills. A Bayesian model averaging approach was taken to identify the best candidate models based upon the hypothesized predictors. Findings showed the strongest support for pretest abilities and initial response to instruction as best predictors of end-of-year outcomes. R[superscript 2] values showed that the most probable models accounted for a substantial amount of variance in student outcomes for Expressive Vocabulary, Uppercase Letter Naming, Letter Sounds, and Concepts About Print. However, none of the candidate models emerged as a substantially better account of the variability in outcomes, suggesting predictors appeared to comparably account for outcomes through multiple combinations. Therefore, while the models explained a substantial portion of the variance for some CBMs, there was no single set of predictors that clearly outperformed others. These findings suggest that responsiveness to TELL was comparably accounted for by various combinations of these best candidate predictors. [The dissertation citations contained here are published with the permission of ProQuest LLC. Further reproduction is prohibited without permission. Copies of dissertations may be obtained by Telephone (800) 1-800-521-0600. Web page: http://bibliotheek.ehb.be:2222/en-US/products/dissertations/individuals.shtml.]