Summary of Findings from Contrasting Groups and Two Rounds of ELP/ELA Analyses

The results from the Contrasting Groups study and the first round of ELP/ ELA analyses conducted in fall 2018 suggested that raising the threshold scores, particularly the threshold demarcating ELPAC Overall PL 4, would allow the ELPAC PL 4 threshold to more closely approximate the PL estimations of EL students given by teachers and the achievement distribution of the monolingual EO students on the state ELA assessment. As a result, policymakers approved new threshold scores for the summative ELP assessment in November 2018. This decision triggered a second round of ELP/ELA analyses examining the achievement of EL students on the state’s ELA assessment disaggregated by the Summative ELPAC performance levels using the newly adopted threshold scores (i.e., 2019 threshold). The results from the second round of analyses suggested that using these threshold scores would substantially reduce the percentage of EL students classified as ELPAC PL 4, which could result in fewer students exiting EL status. Subsequently, additional analyses were conducted to evaluate performance for subgroups of students on the state’s ELA assessment; these are described below.

Comparing Performance on the State's ELA Assessment of EL students with ELPAC Overall PL 4 to Similarly Situated EO Students and New RFEP Students

Given that in several grades EL students meeting the higher ELPAC PL 4 performance threshold neither had an equal likelihood of scoring proficient on the ELA assessment nor approximated the performance of comparison groups (see Figures in 9.5 and 9.6, discussed below), we compared their performance to that of potentially more comparable subgroups within the EO and RFEP groups. Given that ELPAC PL 4 was considered a reclassification criterion, special attention was paid to this level for further analysis. Specifically, using the 2019 thresholds, we compared the performance of EL students scoring at ELPAC overall PL 4 to three subgroups of non-EL students on the state ELA assessment: (a) new RFEP students who were most recently reclassified (in school year 2016-2017) using locally determined district criteria in place across the state’s LEAs; (b) EO students who are similarly situated to the EL student population with respect to economic disadvantage and disability status (that is,

EO students having both “economically disadvantaged” and “non-disability” status.); and (c) similarly situated new RFEP students (i.e., economically disadvantaged without disability status). Regarding similarly situated populations, the rationale for conducting these additional analyses was to consider factors (e.g., socioeconomic status, disability status) that are well documented to be negatively related to academic performance beyond ELP. Regarding new RFEP students, the aim was to consider those students who most recently exited EL status and are therefore most comparable to EL students that meet any proposed ELPAC exit criterion. For example, empirical research (e.g., Thompson, 2017; Umansky, 2016a, 2016b) in California and other states has shown that the longer students have been reclassified, the more likely they are to have received access to grade-level content instruction.

The results are illustrated in Figures 9.5a through 9.5c for the performance of EL students on the state’s ELA assessments at ELPAC Overall PL 3 and PL 4 compared to (all) EO students and new RFEP students at Grades 4, 7, and 11, using the 2019 Summative ELPAC thresholds. Importantly, these figures depict the performance threshold for the state’s ELA assessment Level 3 (with a solid line), as well as that for the midpoint of the scale score range for the state’s ELA Level 2 (Standard Nearly Met) with a dashed line.

The results indicate that for all grades examined (except Grades 4 in Figure 9.5a and Grade 5 not shown), EL students at Overall PL 4 performed very similarly to new RFEP students on ELA. Specifically, they attained or exceeded an equal likelihood of scoring at ELA Level 3 in Grade 3 (not shown) and of scoring at the midpoint of ELA Level 2 in Grades 6 (not shown), 7 (Figure 9.5b), 8 (not shown), and 11 (Figure 9.5c). In Grade 5 (not shown), these EL students performed lower than EO and new RFEP students but still exceeded an equal likelihood of scoring at the midpoint of ELA Level 2. These results provided some support for the Overall PL 4 as the criterion for reclassification.

We repeated the above analysis using from each language group a subgroup of students that had both “economically disadvantaged” and “non-disability” status. The number and percent of each examined subgroup in Grades 3-8 and 11 matched data set satisfying these two “similarly situated” criteria are presented in Table 9.5.

When these similarly situated groups of EL, EO, and new RFEP students are compared, the performance gaps diminish between EL students at ELPAC PL 4 and new RFEP students, compared to EO students. Specifically, the performance of similarly situated EO students declines at all grade levels, while that of EL student and new RFEP students remains the same at Grades 3, 6, 8, and 11 and diminishes only slightly at Grades 4, 5, and 7. The results for students in Grades 4, 7, and 11 are illustrated in Figures 9.6a through 9.6c, respectively. These results provided additional support for using the Overall PL 4 as the threshold for reclassification.

FIGURE 9.5a ELA performance by ELPAC overall performance level and EL status (EO and new RFEP) at Grade 4 (2019 thresholds).

FIGURE 9.5b ELA performance by ELPAC overall performance level and EL status (EO and new RFEP) at Grade 7 (2019 thresholds).

FIGURE 9.5c ELA performance by ELPAC overall performance level and EL status (EO and new RFEP) at Grade 11 (2019 thresholds).

FIGURE 9.6a ELA performance by ELPAC overall performance level and EL status (similarly situated EO and similarly situated new RFEP) at Grade 4 (2019 thresholds).

FIGURE 9.6b ELA performance by ELPAC overall performance level and EL status (similarly situated EO and similarly situated new RFEP) at Grade 7 (2019 thresholds).

FIGURE 9.6c ELA performance by ELPAC overall performance level and EL status (similarly situated EO and similarly situated new RFEP) at Grade 11 (2019 thresholds).

TABLE 9.5 Number and Percentage of Economically Disadvantaged and Nondisabled Students among Each Subgroup, Grades 3—8 and 11 (2019 thresholds).