|HIGH SCHOOL LONGITUDINAL SAMPLE SURVEY:|
|HSLS:09 collects data from:|
The High School Longitudinal Study of 2009 (HSLS:09) is the fifth study undertaken by the National Center for Education Statistics (NCES) as part of its Secondary Longitudinal Studies Program. The predecessors of HSLS:09 include three completed studies (the National Longitudinal Study of the High School Class of 1972 [NLS:72]–see the NLS:72 chapter; the High School and Beyond Longitudinal Study [HS&B]–see the HS&B chapter; and the National Education Longitudinal Study of 1988 [NELS:88]–see the NELS:88 chapter) as well as one ongoing study (the Education Longitudinal Study of 2002 [ELS:2002]–see the ELS:2002 chapter).
HSLS:09 is a nationally representative, longitudinal study of more than 21,000 ninth–graders in 944 schools who will be followed through their secondary and postsecondary years. The study focuses on understanding studentsí plans and trajectories from the beginning of high school into postsecondary education, the workforce, and other early adulthood transitions
What students decide to pursue when, why, and how are crucial questions for HSLS:09, especially, but not solely, with regard to science, technology, engineering, and math (STEM) courses, majors, and careers. Additionally, this study features a new student assessment in algebraic skills, reasoning, and problem solving while, as in past studies, including surveys of students, their parents, math and science teachers, and school administrators, as well as a new survey of school counselors. The HSLS:09 base-year data collection took place in fall 2009 and generated a set of nationally representative data as well as state-level representative data for 10 states.
The first follow-up took place in the spring of 2012, and further depicts the circumstances and implications for later outcomes of process data on student decision–making. With the advent of first follow–up data, HSLS:09 can now measure mathematics achievement gains in the first 3 years of high school. Generally, across both the base year and first follow–up, the study questions students on when, why, and how they make decisions about high school courses and postsecondary options, including what factors, from parental input to considerations of financial aid for postsecondary education, enter into these decisions. Because the study started with fall ninth–graders, it was able to identify high school dropouts in the first follow–up. The antecedent data from the base year will enable researchers to study the process of school disengagement, and will include relatively ďearlyĒ dropout, those who left as early as spring of ninth grade.
The 2013 Update and the High School Transcript Collection of HSLS:09 allow researchers to begin analyses of high school outcomes and immediate postsecondary plans and experiences, bolstered by the rich data gathered in prior rounds. The goal of the 2013 update was to efficiently collect information on sample membersí status with respect to high school completion, postsecondary applications and enrollment, financial aid applications and offers, and employment.
The second follow–up data collection was conducted from March 2016 through January 2017, approximately 3 years after high school graduation for most of the cohort. The data collected allow researchers to examine an array of young-adulthood outcomes among fall 2009 ninth–graders, including delayed high school completion, postsecondary enrollment, early postsecondary persistence and attainment, labor market experiences, family formation, and family financial support. Analyses of these outcomes can capitalize on the large amounts of data gathered about the students in fall 2009, in 2012 (when most were spring–term 11th–graders), and in summer and fall 2013 (when most had completed high school). Analyses can also be augmented with information collected from parents, teachers, administrators, counselors, and high school transcripts.
The core research questions for HSLS:09 explore studentsí secondary to postsecondary transition plans and the evolution of those plans. Included is consideration of paths into and out of STEM courses and the educational and social experiences that affect these shifts. In this regard, HSLS:09 addresses many of the same educational and occupational issues as its predecessor longitudinal studies; however, HSLS:09 places added emphasis on the paths that lead students to pursue and persist in courses and careers in STEM fields.
The student is the fundamental unit of analysis in HSLS:09. In the base–year survey, data from studentsí school, classroom, and home environment were collected and attached to the student record to provide information on the contextual factors that might influence their motivation, beliefs, and interests in goal setting and decision-making. Contextual information was provided by several sources, including the schoolís head administrator, the lead counselor (or staff member most knowledgeable about the entering ninth–grade class), studentsí mathematics and science teachers, and a parent.
Base–Year Survey. The base-year survey was conducted in fall 2009, and included the student questionnaire, the student assessment of algebraic reasoning, and the parent, teacher, school administrator, and counselor questionnaires as described below.
First follow–up survey. The first follow–up questionnaires comprised measures repeated from the base year, in order to measure change in a base–year construct (e.g., educational expectations) or outcome measures (e.g., dropping out of high school) that can be related to base–year antecedents, and augmented by further items that are specific to the first follow–up (e.g., transition to high school ceases to be an emphasis in the first follow–up, but transition plans for postsecondary education loom larger). Instruments were developed, and revised, based on results from the base–year and first follow–up field tests, cognitive interviews, and feedback from the Technical Review Panel (TRP) and Office of Management and Budget (OMB). The contents of the four first follow–up questionnaires–student, parent, administrator, and counselor–are described below. Certain items were deemed ďcriticalĒ (i.e., of special importance to the study), and respondents who skipped such items were prompted, with a message noting the importance of the item and requesting that they provide an answer if at all possible.
2013 Update survey. The 2013 Update survey took place between June and January 2014. The survey, which could be completed by either the sample member or a parent, was designed to gather basic information about the sample memberís high school completion status or plans, postsecondary education and work plans, and the college application and financing process. Questions were adapted so that parent respondents were providing information about their childís activities and plans, although some subjective questions (e.g., about the reasons for choosing a college) were also asked and may systematically differ across parents and children. Because the survey was administered over a number of months that preceded and overlapped with the traditional start of college classes in the fall, a number of questions were anchored to November 1, 2013, and designed to elicit the anticipated or realized activities of respondents on that date.
High School Transcript Collection. High school transcripts were collected in the 2013–14 academic year; methodology was tested in the transcript field test (see the appended field test report, appendix A). Records matching (for example, ACT and SAT scores, Free Application for Federal Student Aid [FAFSA] data, General Educational Development [GED] data) also contributed to the dataset.
Between fall 2013 and spring 2014, high school transcripts were gathered from all schools that students had attendedóincluding schools known from prior rounds, schools identified by the student or parent during the 2013 update survey, or schools discovered during the request for transcripts from existing schools. Coursetaking records from transcripts were keyed and coded using the School Courses for the Exchange of Data (SCED), a code frame that classifies courses into standard 12-digit codes reflecting their course content and placement within subjects. HSLS:09ís use of SCED marks the first implementation of this code frame on any NCES transcript study. Course credits, course grades, and other measures derived from transcripts were standardized to ensure comparability across schools. In addition, student records were matched to external data sources to obtain data on SAT and ACT scores, Free Application for Federal Student Aid (FAFSA) data, and GED completion data.
Second follow-up survey. The second follow-up was designed to collect information from the cohort approximately 3 years after the modal high school completion date. The second follow-up survey explored a variety of academic and employment-related topics that include, but are not limited to, high school completion and experiences, college enrollment history and future enrollment plans, and employment and unemployment history. The second follow-up survey collected information on each of these diverse activities while continuing to capture information on studentsí experiences and the influences and constraints on their decision-making about postsecondary education, fields of study, and occupations. Selected survey items paid particular attention to studentsí experience with and choices related to STEM fields. Respondents were asked to anchor their responses for a number of questions to the end of February 2016 (the month before data collection began), and many questions asked specifically about the respondentís activities and status in that month.
The second follow-up also covered a range of topics related to family, community, and personal characteristics including marital and parental status, household composition, financial well-being, community engagement (e.g., citizenship, voting registration, volunteering), personal characteristics and values (e.g., disabilities, sexual orientation and gender identity, experiences of discrimination, life values), and significant life events (e.g., job loss, death, serious injury or illness).
Student financial aid and postsecondary transcript collection. In addition to information obtained from sample members who participated in the second follow-up survey, data were obtained in 2017 from postsecondary institutions as part of the student financial aid records and postsecondary transcript collection. Financial aid data records include data collected from the institutions that sample members attended, and federal student loan records from the Department of Educationís Office of Federal Student Aid. Records collected from the institutions attended by HSLS:09 sample members provide detailed information about studentsí enrollment patterns, degree programs or other programs of study, progress toward degree, and costs of attendance. The postsecondary transcript data cover postsecondary coursetaking through 2016 and provide detailed information on studentsí academic experience, including academic performance, credit accumulation, enrollment periods, and transfer between institutions.
Student questionnaire. The student questionnaire was typically self–administered using a computer during in–school interview sessions. If a student was unable to participate in school, a telephone interview was conducted using the same survey instrument with the addition of interviewer instructions.
Background information was collected from the ninth–graders, including demographic information (such as sex, race/ethnicity, birth date, and native language) and the names, addresses, and phone numbers of people who would know how to locate them for future rounds of the study. Information was also gathered on studentsí recent school experiences, including the school they attended in the previous school year (2008–09) and their grade level at the time; their involvement with various math and science activities since the beginning of the previous school year; and the math and science courses they took in the eighth grade, along with the final grade earned in each.
Questions were also asked about studentsí self–efficacy in math and science and self–identification as a math and science person. Additionally, data were collected on the math and science courses they were taking in the fall of 2009, and students were asked to identify the teachers of these courses. Questions were asked about studentsí attitudes toward school, math, and science, as well as about whom students spoke with regarding their future education, career plans, and personal problems. Moreover, data were collected on friendsí attitudes about school and related behaviors, as well as on programs in which the student had participated, such as Upward Bound or MESA (Mathematics, Engineering, Science Achievement). Students were also asked about their perceptions of malesí and femalesí abilities in math, science, and English and language arts. Finally, information was collected on their high school, career, and college plans. Specifically, students were asked about their plans to take additional math and science courses in high school, career or college plans, plans to take standardized college placement exams, and general plans for the year after high school. Data were also collected on studentsí educational expectations, such as how confident they were of graduating from high school, as well as their estimates of the cost of college and their expected occupation at age 30.
The first follow–up Student Questionnaire targeted the fall 2009 ninth–grade cohort members in the spring term of the 2011–12 school year, regardless of their school enrollment status (i.e., whether they are students, dropouts, or early graduates). The questionnaire was designed with content appropriate for dropouts and early graduates, as well as students still enrolled in the base–year school, those who have left the base–year school for homeschooling, or those who have transferred to a new school. The first follow–up student questionnaire was a web survey.
Some first follow–up participants were nonrespondents in the base year. Therefore, a number of questions were asked only of sample members for whom the information was missing in the base year. These items pertain to critical classification variables such as language use, and parental education and occupation.
The goal for developing the content of the second follow-up instrument was to design a web survey that, in conjunction with data collected in previous rounds of the study, could provide information on sample memberís choices related to secondary education completion, postsecondary education, entry into the workforce, and family formation. The second follow-up survey not only collected new information about sample membersí activities, but also updated information obtained in previous data collections and gathered information that was missing in prior rounds of the study (either from item- or unit-level nonresponse).
The second follow-up survey was designed for self-administration on the Web as well as interviewer administration via telephone or in-person interviews. The web instrument could be completed on a desktop or laptop computer or a mobile device such as a tablet or smartphone. The survey did not have to be completed in a single session or in a single mode. Respondents could begin the survey, log out, and resume where they left off at a later time. Moreover, respondents could switch modes across survey sessions. For example, they could begin the survey on their own and later complete the survey with an interviewer. For sample members who decided to complete the survey over the phone or in-person, trained interviewers administered the survey, following instructions on each screen that indicated how each question was to be administered (e.g., whether the response options were to be read aloud, and when to probe for more information).
In developing the second follow-up survey instrument, a primary challenge was to minimize respondent burden while collecting as much information as possible. One way to do this was by routing the survey around questions when the answer could be logically inferred from the answer provided to an earlier question. Additionally, the full-length survey instrument was used from the beginning of data collection through December 11, 2016, at which time it was replaced by an abbreviated instrument to encourage the remaining nonrespondents to participate. Survey respondents who partially completed the full-length instrument before that date and resumed after that date were automatically switched from the full-length to the abbreviated survey.
Student assessment of algebraic reasoning. The HSLS:09 assessment battery was designed to allow for comparison of students’ algebraic reasoning skills across years. Algebraic reasoning skills were first assessed in the fall of 2009, when students were in the 9th grade, and were assessed again in the spring of 2012, when most students were in the 11th grade.
The assessment was designed to measure students’ understanding, and growth in understanding, of key algebraic knowledge and skills useful in the preparation for the study of science, further study within the mathematical sciences and statistics, and the requisite skills expected in the workplace. Accordingly, the framework was designed to assess a cross–section of topics representative of six major content domains of algebra (the language of algebra; proportional relationships and change; linear equations, inequalities, and functions; nonlinear equations, inequalities, and functions; systems of equations; and sequences and recursive relationships) and four key processes of algebra (demonstrating algebraic skills; using representations of algebraic ideas; performing algebraic reasoning; and solving algebraic problems).
As with the base year, the HSLS:09 first follow–up mathematics assessment was administered by computer, using a two–stage design wherein each student completed a Stage 1 “router test” and then a Stage 2 test designated as “low,” “moderate,” or “high” difficulty that was assigned on the basis of Stage 1 performance. The first follow–up assessment consisted of 73 unique items, with 23 serving as linking items to the base–year assessment, with any given student receiving 40 items. The computer–delivered design included an online scientific calculator and allowed students to skip and return to items within each stage and to identify items for review within each stage before submitting their answers as finished.
Parent questionnaire. The parent questionnaire was completed by the parent or guardian most familiar with the ninth–grader’s school situation and experience. Most often this was the student’s mother or father (although, in rare instances, a guardian such as a grandparent responded). The questionnaire collected information on the presence of parents or guardians in the household, their relationship to the ninth–grader, and their marital status; the parents’ race and ethnicity, immigration status, language use, and socioeconomic status; the student’s place of birth, immigration to the United States and grade placement upon arrival (if born abroad), and whether the student had ever been or was currently enrolled in a program for English language learners; the student’s educational history (e.g., skipping or repeating grades, changing schools, dropout episodes, suspensions and expulsions, special education services, enrollment in honors courses); the parents’ involvement in the ninth–grader’s education and learning; the parents’ plans and preparations for their child’s postsecondary education; and contact information for parents, relatives, and friends who could locate the ninth–grader in subsequent rounds of the study.
In the first follow–up, a random subsample of students’ parents were administered the parent questionnaire. Data collection staff asked that the parent or guardian most familiar with the school situation and experience of the student sample member complete the parent questionnaire. As with the student questionnaire, there are questions that have been adapted to the situation of parents of dropouts as well as to parents of school attendees. The first follow–up parent questionnaire was fielded as both a web survey and a computer–assisted telephone or in–person survey. In the beginning of data collection, the survey was fielded with the web option. As the period of data collection elapsed, parents were also given computer–assisted options, with an interviewer over the telephone and at the end of the field period an in–person interviewer.
Parents were asked about their relationship to their child, how much of the time the cohort member lives with the parent respondent, if other parents reside in the household, parent respondentís current marital status, counts of household members by age, school enrollment status of the student sample member, negative life events, prior educational experience including grades their child repeated, school suspensions, dropout episodes, number of times parent contacted the school, family activities, parent-child activities to prepare for the postsecondary transition, parent aspirations and expectations, ability to complete a bachelorís degree, ranking of importance of various college features, degree of parent and student input for postsecondary decision-making, affordability of college, means of getting financial aid information, expectations for qualification for financial aid, obstacles to applying for financial aid, savings for education, willingness to borrow, family educational and occupational background, employment status, income, demographic background, and languages spoken in the home. Some questions (e.g., national origin) that were asked in the base year are only repeated in the first follow-up if base-year data are missing due to unit or item nonresponse.
Teacher questionnaire. All teachers who had an HSLS:09 student in their math or science course were eligible for the teacher questionnaire. Teachers were asked to answer questions regarding their own demographic characteristics, educational history, certification, and teaching history. Additionally, they were asked to evaluate their departments, departmental colleagues, principal and faculty.
As well, the teacher questionnaire included questions about students in general (i.e., they were not asked questions about any particular students). For instance, teachers were asked to report on their beliefs about the influence of a student’s home environment on their ability to be effective teachers; how male and female math and science abilities compared; and how they assessed the achievement levels and preparedness of students in their class. As well, teachers provided information on the use of small groups in class and their emphasis on various course objectives.
Importantly, the teacher data are meant only to supply contextual information for students’ classrooms, while the student remains the unit of analysis. The teacher sample is not representative of teachers in the school. The design of this component does not provide data for a standalone analysis sample of teachers, but instead permits specific teacher characteristics and practices to be related directly to the learning context and educational outcomes of sampled students.
School administrator questionnaire. Most of the school administrator questionnaire could be completed by the principal or another knowledgeable staff member at the school. Respondents were asked factual questions about the schoolís characteristics, including grade span, control (public or private), type (e.g., charter, magnet, single sex, religious), academic calendar, and course scheduling. Questions were also asked about the student body, school faculty, and math and science curriculum. The final section, which could only be answered by the school administrator, included questions about the administratorís background and an evaluation of the schoolís problems and challenges.
In the first follow-up, the school administrator questionnaire targeted the base-year schools, 2.5 years later. In addition, an abbreviated version of the administrator survey was fielded to collect information from schools to which students in the study transferred. The school administrator questionnaire was fielded as a web survey.
The full administrator survey consisted of four sections: (1) school characteristics; (2) programs, policies, and statistics of the school; (3) school staffing; and (4) opinions and background of the school principal. The school characteristics section contains questions about the school type (e.g., regular, charter, alternative); magnet and school choice programs; academic calendar; course scheduling; hours of instruction; and percentage of students who attend area or regional career and technical schools. The second section includes questions about enrollment; proportion of students who receive free or reduced–price lunch, who are English language learners, and who receive special education services; enrollment and assignment policies; average daily attendance; absenteeism policies; programs to help students who are struggling academically; credit recovery programs; alternative and dropout prevention programs; activities to increase student interest and achievement in math and science; and years of coursework required for graduation. The third section asks questions about teachers within the school. Topics include numbers of teachers by full– and part–time status and by subject matter; teacher recruitment and retention; teacher absenteeism rates; and support for new math and science teachers. The fourth and final section includes questions on counseling goals and emphases; difficulty and methods of filling teaching vacancies for science and math; principalís perception of school problems; principal demographic characteristics; and principalís educational background and experience.
Because the first three sections contained factual questions about the school, these questions could be completed by the principal or another knowledgeable individual designated by the principal. However, the final section contains background and subjective questions, and the only appropriate respondent is the principal. Therefore, different login credentials were issued to school administrators and their designees such that school administrators were able to access the entire questionnaire, while designees were able to access only the first three parts. In an effort to reduce the burden of reporting detailed statistics, respondents were instructed that informed estimates were acceptable.
In addition to the full school administrator questionnaire, an abbreviated version was sent to schools to which students had transferred. The abbreviated version could be completed by a knowledgeable person in the school administrator’s office by web, computer–assisted telephone interview (CATI), or paper–and–pencil interview (PAPI) instruments. The abbreviated version included a subset of questions. These include questions about the school type (e.g., regular, charter, alternative); hours of instruction; course scheduling; enrollment; proportion of students who receive free or reduced–price lunch, who are English language learners, and who receive special education services; postsecondary destinations of seniors; numbers of full– and part–time teachers, and math and science teachers; and years of service of the principal.
Counselor questionnaire. The counselor questionnaire was filled out by the lead counselor (or staff member most knowledgeable about the entering ninth–grade class) at each school, and gathered information on the total number of full–and part–time counselors at the school, the number certified as high school counselors, and the average caseload per counselor. Information was also collected on the school’s counseling program and tasks performed by the counselor. The questionnaire also focused on how counselors and the school as a whole assisted eighth–grade students’ transition into ninth grade and the school’s use of career and education plans. Counselors were asked to remark on programs and services offered to students, such as enrichment courses, assistance for struggling students, dropout prevention programs, encouragement of the pursuit of math and science education and employment, and assistance with the transition from high school to college or the workforce. Counselors also reported on the criteria used to place ninth–graders and upperclassmen in math and science courses. In addition, background information on the school counselor, including how he or she entered the profession, how many years served as a counselor, and educational history, were collected. Since the head counselor at each school was asked to complete the questionnaire, the respondents do not constitute a standalone nationally representative sample of high school counselors (or 9th–grade counselors).
As in the base year, in the first follow–up the head or senior–most counselor at each base–year school was asked to complete the survey. The resulting counselor data are purely contextual, linked to the basic unit of analysis, the student sample member. The student, in turn, will have no first follow-up counselor data if she or he transferred to a new school, went into homeschooling, or attended a base–year school in which the counselor did not participate in 2012. The school counselor questionnaire was fielded as a web survey.
The counselor survey contained four sections: (1) counselor staffing and practices, (2) programs and support for students, (3) math and science placement, and (4) school reporting and statistics on students. The first section includes questions on number of full– and part–time counselors, average caseload, method of assignment to students, breakdowns of percentage of time spent between delivering various services to students, and counselor duties and functions.
The second section contains questions such as programs and supports offered by the school, dual or concurrent enrollment offerings, summer enrichment, sources of credits beyond those offered directly by the school, attention given students in need of extra assistance, dropout prevention programs and services it offers, General Educational Development preparation, assistance with college entrance exams, assistance identifying and applying to colleges and universities, modes of assistance with college or university applications or financial aid and Free Application for Federal Student Aid preparation, programs and initiatives to ease the transition from high school to work, percentage of juniors and seniors taking advantage of various work preparation services, and school linkages with local employers.
The third section includes questions such as factors associated with mathematics and science course placement and sequencing, importance of various factors for advanced science and math placement, onsite and offsite calculus and physics, student participation and success in Advanced Placement and International Baccalaureate courses and exams, and average SAT and ACT scores of the school. The last section contains questions about the types of transition and outcomes data collected and analyzed by the school.
The base–year data collection of HSLS:09 took place in the fall of the 2009–10 school year. The first follow–up took place in the spring of 2012, when most sample members were in the 11th grade. A postsecondary update took place in the summer/fall of 2013, and high school transcripts were collected in the 2013–14 school year (i.e. fall of 2013 and spring of 2014). The second follow–up took place March 2016, through January 2017, approximately 3 years after high school graduation for most of the cohort. The number and timing of future follow–ups is yet to be determined, although the expectation is that the cohort will be followed at least to age 30, with a questionnaire administration and a postsecondary education transcript collection in 2025–26.
Public-use data for HSLS through the second follow-up data collection can be obtained at https://nces.ed.gov/onlinecodebook. Information on restricted-use data is available at https://nces.ed.gov/pubsearch/getpubcats.asp?sid=111. Additional data on HSLS is projected to be available in Summer 2019.