Springfield is a case of a district where implementation was strong, and where the sample size of the MathForward™ group was large enough detect statistically significant effects, so the results are not due to chance. At the same time, the differences between gains in 2007 to 2008 may have been caused by factors other than MathForward™. In addition, instability in estimates of gain scores from year to year make it difficult to determine whether measurement error or differences in test difficulty from year to year are the cause of the observed change.