• No results found

Populations and Variables (Quantitative Data)

Appendix: Data, Methods, and Limitations

2. Populations and Variables (Quantitative Data)

2.1 Populations Compared

2.1.1 Final Report. The following figure describes the groups analyzed in the aggregate report. When grouped by their native language, students are classified as native English speakers (NES) or native speakers of other languages (NSOL). When grouped by language proficiency, native speakers of languages other than English are grouped into those who are English proficient (EP) and those who have limited English proficiency (LEP). The term “LEP” is defined by the Massachusetts Department of Elementary and Secondary Education as “not able to perform ordinary classwork in English.” When grouped by program participation, students are either in programs for English Learners (In EL Programs) or in General Educa- tion (In General Ed). Enrollment, demographic, and outcomes analyses are conducted for each group, by native language, language proficiency, and program participation. The last row compares all four groups of students: NES in General Ed, NSOL EP in General Ed, LEP in General Ed, and LEPs in EL programs.

Total All BPS

Native Language NES NSOL

Language

Proficiency EP EP LEP

Program Participation of LEPs

In General Education In General Education In Gen Ed In EL Programs

All Program

Participation NES In General Education NSOL EP in General Education LEP in Gen Ed LEP in EL Programs

• NES LEP students: In analyzing LEP students, the dataset included a small number of students who were coded as native English speakers as well as limited English proficient. The number was small and did not change the outcomes. Given the analysis scheme, the decision was made to leave them with the NES EP group. • LEP students in General Ed: LEP students in General Education are students who

have opted out of programs for English Learners or who have transitioned to General Education but still retain their LEP designation.

• Grade levels: For all except the last row and MCAS results, these analyses are presented for all students in each group as well as disaggregated by grade level. Grade level variables were created as follows:

Elementary Grades K0, K1, K2, 1, 2, 3, 4, 5 Middle Grades 6, 7, 8

High Grades 9, 10, 11, 12

For schools that span grade levels, students are analyzed by the grade level they are in. For example, students in a K2–8 school are separated into those in the elementary grades (K2–5) and those in the middle grades (6–8) for the grade level analyses.

2.1.2 Language Groups Report. The language group report includes analyses of the enrollment, program participation, and outcomes of the five largest non-English native language groups: Spanish, Haitian Creole, Vietnamese, Cape Verdean Creole, and Chinese dialects. For the analysis presented in the language group reports, the students who “opted out” of programs for ELs are not included in General Education programs but are part of the total analysis of BPS students. In most tables for the language group reports, we compare the outcomes of EL students from these native language groups with all students in General Education, all students in EL programs, and native speakers of the specific language in General Education. Students who are native speakers of a language other than English but who are in general programs include former students of programs for ELs as well as students who may be native speakers of a language other than English but who were never enrolled in a program for ELs.

2.2 Definition of Demographic and Program Participation Variables (both reports)

2.2.1 Demographic Variables. The demographic variables included in the data set included gender, race/ethnicity, native language, and free/reduced price lunch status. Each analysis using a demographic variable represents the proportion of students in that category in the grouping being analyzed. For example, the proportion of Black students who are native English speakers is calculated as the total number of Black students who are NES divided by the total number of NES students.

• Gender—Students are male or female.

• Race/ethnicity—Students are classified into one of five categories: American Indi- an/Alaska Native; Asian/Pacific Islander; Black; White; or Hispanic. The proportions were calculated for each category. Groups sizes for American Indian/Alaska Native in BPS were too small to report. In AY2006, race/ethnicity reporting categories changed. Variables were created to group students into the original race/ethnicity variables to allow consistent reporting.

• Native language—Students in Boston speak many native languages. The six most frequently cited native languages during the study period were: English, Spanish, Haitian Creole, Vietnamese, Cape Verdean Creole, and Chinese dialects.

• Receiving free/reduced price lunch—Receiving free or reduced price lunch is the most commonly used indicator of the income level of students. In the absence of complete and accurate data on household income, lunch status serves as a proxy for income. The lunch status variable provided was “receiving free lunch” or “receiv- ing reduced price lunch” or “not eligible for either.” The first two categories were collapsed into one: “receiving free or reduced price lunch.” This variable underes- timates the level of poverty in the group, since many students who are eligible by status may not actually receive the service.

Language proficiency. The language proficiency variable, LEP, was an ill-defined and inconsistently used term in the aftermath of Question 2, as described in this study. However, it is the variable that was used to identify students in need of English Learner services and programs and therefore serves to delineate native speakers of other languages who were English proficient from those who were not. Given what we have learned about mis-assessment, family under-reporting, and lack of common understanding among BPS staff, the LEP variable is an undercount. There are likely to be NSOL EP students in General Education programs who are actually LEP.

Participation in programs for English Learners. The program participation variables in the data set were re-coded such that LEP students in any program for English Learners were “In EL program” and LEP students who were not in a program were “In General Ed.” Students who were coded as “waiver” or “opt out” were re-coded as “In General Ed.”

Participation in Special Education programs. Students are classified into 13 categories in the BPS data set. To reflect predominant practices and simplify report- ing, a new variable was created that combined the classifications into three groups: not in Special Education, in partial or full inclusion Special Education, and in substantially separate Special Education.

2.3 Definition of Outcomes Variables (both reports)

Using the research literature on effective schools and student outcomes, as well as the avail- able data from the BPS data set, the following outcomes identified to be studied: attendance, suspension, transfer, drop-out, grade retention, and MCAS pass rates. All outcomes variables, except for attendance rates, represent the proportion of students within each school type who reflect that measure. For example, in terms of out-of-school suspensions, the numbers represent the percentage of students within each school type who were suspended in each school year. Median attendance rates were calculated as the days students attended divided by days of membership each year; the median rate is reported.

Outcome variables studied in the language group report mirrored the ones in the aggregate report. The only exception was MCAS outcomes, for which the numbers of test takers from the language groups except Spanish speakers are too small to report.

Variable Definition

Attendance rate Median percentage of days that students attended school in a given year divided by days of membership for each school type each year.

Out-of-school suspension rate The proportion of students who were suspended from school at least once in each school year.

Transfer rate The proportion of students who transferred out of the district in a given school year. This includes students who transferred to any school outside of the Boston Public Schools as well as students who dropped out of school without official notification.

Annual drop-out rate The proportion of students who dropped out of the district in a given school year. This indicator does not exclude 0/1s.

Grade retention rate The proportion of students in a given school year who were not pro- moted to the next grade.

MCAS English Language Arts

and Math pass rate The sum of the proportions of students scoring in the advanced, profi-cient, and needs improvement performance categories on the English Language Arts or Math MCAS exams in a given year.