While verbal intelligence was included in the analysis as a covariate,
the results pertaining to this variable are worth noting. There was a
strong positive relation between verbal intelligence and mathematics
achievement at 54 months, b = 8.92, p b .001. In other words, at preschool
age, children with high verbal intelligence showed better math
performance than those with low verbal intelligence. At the same
time, there was a negative interaction between verbal intelligence and
the linear rate of growth in math achievement, b = −1.97, p b .001.
This means that, across elementary school, children who had lower
PLS scores in preschool began catching up in math with those who
had higher PLS scores. Thus, the predictive relation of verbal intelligence
to math achievement decreased over time. Finally, therewas an interaction
between PLS scores and the quadratic rate of growth in math
scores, b = 0.17, p b .001, indicating that while the math achievement
gap associated with verbal intelligence diminished over time, there
was a deceleration in the rate at which it diminished.