Skip Navigation
Funding Opportunities | Search Funded Research Grants and Contracts

IES Grant

Title: Project PIMSELA: Partnering to Investigate Math and Science English Learners' Access and Achievement
Center: NCER Year: 2017
Principal Investigator: Varghese, Manka Awardee: University of Washington
Program: Researcher-Practitioner Partnerships in Education Research      [Program Details]
Award Period: 2 years (9/1/2017-08/31/2019) Award Amount: $397,500
Type: Researcher-Practitioner Partnership Award Number: R305H170019

Co-Principal Investigators: Lee, Min; Sanders, Elizabeth; Gallardo, Veronica; Anderson, Eric

Partner: Seattle Public Schools

Education Issue: The central aim of the "Partnering to Investigate Math and Science English Learners' Access and Achievement" (PIMSELA) project was to understand underlying factors that explain English learner (EL) students' diminished access to, and underachievement in, middle and high school math and science in Seattle Public Schools (SPS). The results of our collaboration allowed PIMSELA partners to consider policy recommendations for improving equitable EL participation in math and science during middle and high school, as well as improving equitable EL high school graduation rates and access to postsecondary education.

Partnership Goals: Project PIMSELA was focused on collecting, cleaning, coding, and analyzing data related to EL and non-EL course-taking patterns in middle and high school math and science through a university partnership with the SPS district. To this end, the district partner gathered existing administrative data from multiple sources, and the university partner collected primary quantitative survey data as well as qualitative case study data at two schools. Although these data sources have been analyzed, an ongoing goal continues to be integrating the disparate data sources into combined files. In addition to analyzing and integrating data sources, the partnership goal was also to disseminate products and policy recommendations based on the research findings, and to develop a future research plan; these latter goals are still in progress, and the team continues to work toward their fulfillment during 2021.

Partners and Partnership Activities: The PIMSELA project is a policy-oriented partnership between the University of Washington's College of Education and the SPS District. The structure of the PIMSELA partnership included a lead team of researchers, educators, and graduate students, all of whom engaged collaboratively in research and development activities of groundwork, theorizing and sense making, dissemination, and policy recommendations. Within the first year of the project, the partnership itself evolved with an expansion of the team from two SPS district departments to five. The partners actively met for two years, both regularly and for ad hoc activities. Additionally, the partnership team collectively reflected on the partnership using both a survey (anonymous) as well as in-person dialogue. The team plans to submit an exploratory grant proposal based on what has been learned during those activities.

Setting: The PIMSELA partnership collected and used data from SPS, the largest and most linguistically diverse urban public school district in Washington State. Thirteen percent of SPS's total student population has been identified as EL, representing over 132 home languages. The university team is located within the same city and some of the partners had already worked together on previous collaborations. 

Population/Sample: The project samples were collected from multiple sources. Below is information on the population/sample for each data source.

ERDC data source. In the first phase of the quantitative analyses, the team included the data Available from Washington State's Education Research & Data Center (ERDC) regarding high school math course taking completion and timeliness. The team specifically examined "ever" English Language Learner (ever-ELL; N = 4,984) compared with "never" ELL (N = 28,992) patterns from 2004-05 through 2012-2013 (i.e., the data for students who would have had the opportunity to graduate before the project started in 2017).

SPS 2016 cohort data source. The 2016 cohort included 4,551 students from 21 middle and 11 high schools; of these, 947 (19%) were classified as ever-ELL based on the presence of ELL test data. For these data, the team first examined administrative records linking grades 6—9 advanced math and science course taking with college enrollment. For the same cohort, the team also examined ELL disparities and high school science course taking (grades 9-12) to understand course taking outcomes and patterns of reclassified (or "exited") English learner (ELL) students that included three mutually exclusive student subgroups: Reclassified ELs (N = 362), Still ELLs (N = 227), and Never ELLs (N = 1,888). (Note: the analytic sample was slightly smaller than the complete cohort as some students were missing science course-taking information.)

District high school teacher and counselor staff survey data source. The team created and implemented a novel district-wide survey of all high school math, science, and ELL teachers, as well as counselors, to understand perspectives and understandings of roles and responsibilities in math and science course taking for ELLs. Of the 321 participants invited from 11 schools across five regions, N = 82 (26%) responded. For math items, the sample included n = 55 participants (37 math teachers, 10 counselors, and 8 ELL teachers); for science, the sample included n = 48 (26 science teachers, 10 counselors, and 12 ELL teachers). Items included both quantitative and qualitative data.

High school case studies data source. In the two purposively selected case study schools, the team interviewed 9 teachers, 2 counselors, 2 principals, and 46 students. These data were transcribed for qualitative analyses.

Initial Studies and Analyses:  Both quantitative and qualitative data analyses were employed by the university research partners. Quantitative analyses included descriptive statistics as well as traditional regression and random effects/multilevel models using R, SPSS, and HLM. Qualitative analyses included deductive and inductive coding.

Publications and Products

Policy Brief

Sun, M., Anderson E., and Bastian, K. (2019). Using Data in Evidence Based Policy Processes through Building Research-Practice Partnerships. Education Policy Analytics Lab.