scores The results for models of EOWPVT scores are also shown in Table 5. In the model that included the time teachers spent before,during, and after reading, duration dedicated to after-reading inter-actions was a significant predictor of posttest EOWPVT scores( = 0.022, p = .032). Thus, a 1-s increase in time after reading was predictive of a .022-point increase in posttest EOWPVT scores.Frequency of vocabulary-association questioning was also a significant positive predictor ( = 0.305, p = .019). Each additional instance of a teacher associating vocabulary was predictive of a 0.305-point increase in posttest EOWPVT scores. Finally, duration of vocabulary-association questioning was a significant positive predictor of posttest EOWPVT scores ( = 0.071, p = .041). That is, a1-s increase in time spent associating vocabulary was predictive of a .071-point increase in posttest EOWPVT scores. Posttest EOW-PVT scores were not predicted by comprehension variables, either frequency or duration.In the models that included interaction terms, which were used to test whether the effect of type of talk on vocabulary scores depended on students’ beginning-of-year vocabulary scores, no interaction terms were found to be significant in any of the models for either PPVT or EOWPVT. Therefore, we do not report any results of these models in tables or discuss them in any further detail.