Data sgp adalah suatu tujuan dari SGP. It is used to estimate latent achievement traits and calculate growth percentiles for students in grades 4 through 8, including grade 10 (since there is no MCAS testing in grade 9). SGP data are available for all students with valid test scores on the prior grade level tests, unless otherwise specified by the state. This data is compiled by DESE to provide teachers with information about student growth over time, including a range of student outcomes, for all subjects taught in their classrooms.
Although the aggregation of student performance data has been a significant accomplishment in its own right, there is still a long way to go for researchers and policymakers to fully appreciate the value and potential limitations of this type of measurement. This is especially true if the aggregated SGPs are interpreted as indicators of teacher effectiveness. As the discussion below will show, relationships between aggregated SGPs and student covariates are difficult to interpret, even if we assume that the correlations between latent achievement traits and student background characteristics are causal.
The most important limitation of SGPs as indicators of teacher effectiveness is that they may be biased if used to assess teacher quality. Specifically, the fact that students with similar background characteristics are sorting to schools and teachers of different effectiveness levels can lead to systematically varying aggregate SGPs for these same students. These differences in aggregate SGPs would exist whether or not teachers were able to affect individual students’ achievement through their teaching.
These differences in aggregated SGPs can be eliminated by estimating teacher effects through value-added models that regress student test scores on teacher fixed effects, prior test scores, and student background variables. However, this approach does not remove the relationships between aggregated SGPs and student background variables, which can be seen in the correlations between aggregated SGPs and the covariates listed in Table 3 for math and ELA.
Another possible source of bias in the interpretation of aggregated SGPs is that some covariates have much stronger effects on student achievement than others. For example, students who are absent from school often have lower true SGPs in math and ELA than their peers, which is reflected in the higher within-grade, across-year correlations between these two covariates in Table 3. This type of bias is more difficult to detect in our current analysis because it can occur over many years, making it hard to identify a single cause and effect.
To avoid this problem we recommend that you use the lower level functions (studentGrowthPercentiles and studentGrowthProjections) to produce a set of graphical displays associated with your SGP results. This utility function currently includes facility to produce a variety of charts showing student progress over time as well as interactive bubble charts that display the growth trajectory for a given student. If you are unsure of the specific graphical display you require please contact us with your questions and we will be happy to help.