Title page
Contents
Chapter 1. Introduction 9
1.1. Overview of Data File Documentation (DFD) Report 9
1.2. Historical Background: NCES Secondary Longitudinal Studies Program 10
1.3. High School Longitudinal Study of 2009 13
1.3.1. Base Year, First Follow-up, 2013 Update, and Second Follow-up 14
1.3.2. PETS-SR 16
1.3.3. Research and Policy Issues and Analytic Levels 17
Chapter 2. Sample Design 24
2.1. Base-year Sample Design 24
2.2. First Follow-up Sample Design 25
2.3. 2013 Update and High School Transcript Study Sample Design 26
2.4. Second Follow-up Sample Design 27
2.5. Postsecondary Education Transcript and Student Records Sample Design 27
Chapter 3. Data Collection Methodology and Results 28
3.1. Postsecondary Education Transcripts and Student Records Systems and Processes 28
3.1.1. Postsecondary Data Portal Website 28
3.1.2. Institution Contacting Staff Training 29
3.1.3. Institution Contacting and Recruitment 29
3.2. Postsecondary Education Transcripts-Specific Processes and Quality Control 32
3.2.1. Data Receipt Procedures 32
3.2.2. Transcript Keying/Coding System and Keyer/Coders 33
3.2.3. Coding Taxonomies 35
3.2.4. Transcript Keying/Coding Quality Control 36
3.3. Postsecondary Education Transcripts Data Collection Results 40
3.4. Student Records-Specific Processes and Quality Control 42
3.4.1. Student Records Instrument 42
3.4.2. Student Records Quality-Control Procedures 44
3.5. Student Records Data Collection Results 44
Chapter 4. Data Processing and Editing 47
4.1. Data Processing 47
4.1.1. Transcript Data Reassignment and Consolidation 47
4.1.2. Processing Student Records Data 48
4.2. Data Editing, Documentation, and Review 49
Chapter 5. Response Rates, Analytic Weights, Variance and Design Effects Estimation, Nonresponse Bias Analysis, Imputation, and Disclosure Avoidance 51
5.1. Criteria for Defining Respondents 51
5.2. Unit Response Rates 52
5.3. Overview of Weighting 56
5.3.1. Analysis Weights 56
5.3.2. BRR Weights 63
5.3.3. Weight Characteristics 64
5.3.4. Weighting Quality Control 66
5.4. Choosing an Analytic Weight 67
5.5. Measures of Precision: Standard Errors and Design Effects 78
5.5.1. Standard Errors 78
5.5.2. Design Effects 82
5.6. Unit and Item Nonresponse Bias Analysis 86
5.6.1. Unit Nonresponse Bias Analysis 86
5.6.2. Item Nonresponse Bias Analysis 90
5.7. Single-value Item Imputation 99
5.7.1. Imputed Items 100
5.7.2. Evaluation of the Imputed Values 103
5.8. Disclosure Risk Analysis and Protections 103
5.8.1. PETS-SR Data Products 104
5.8.2. Recoding, Suppression, and Swapping 104
Chapter 6. Data File Contents 106
6.1. PETS-SR Data Products 106
6.1.1. Restricted-use Data Products 106
6.1.2. Public-use Data Products 107
6.2. Contents of the PETS-SR Data Products 108
6.3. Variable Naming Schema 112
6.4. Missing Data 112
6.4.1. Reserve Codes 113
6.4.2. Placeholder Records 113
6.5. Composite Variables 114
6.6. Data Anomalies and Considerations 115
References 117
Table 1. Upcoding of "other, specify" responses: 2018 39
Table 2. Eligible institution participation, by institution type: 2018 41
Table 3. Student-level transcript collection results: 2018 42
Table 4. Number and percent of participating institutions, by student records collection methods, by institution type: 2018 45
Table 5. Student-level student records collection results: 2018 46
Table 6. HSLS:09 unit response rates 55
Table 7. Descriptive characteristics of PETS-SR survey weights: 2018 64
Table 8. Weighted counts and percentages of X2SEX for restricted- and public-use files, by PETS-SR survey weight: 2018 65
Table 9. HSLS:09 analysis weights: 2018 70
Table 10. Number and percentage of completed surveys, high school transcript responses, postsecondary transcript and student records responses, or their combinations... 73
Table 11. Average design effects (deff) and root design effects (deft) for postsecondary transcript and student records variables 85
Table 12. Summary statistics for unit nonresponse bias analyses before and after weight adjustments for nonresponse, by HSLS:09 PETS-SR analysis weights: 2018 88
Table 13. Student records items with a weighted item response rate below 85 percent using SR student weight (W5PSRECORDS) 93
Table 14. Student records items with a weighted item response rate below 85 percent using PETS student weight (W5PSTRANS) 95
Table 15. Frequency distribution of the estimated bias ratios for student records items 96
Table 16. Frequency distribution of the estimated bias ratios for transcript items 97
Table 17. Summary statistics for student records item nonresponse bias analyses using WSPSRECORDS weight 98
Table 18. Summary statistics for student-level item nonresponse bias analyses using W5PSTRANS weight 99
Table 19. Student records variables included in single-value imputation, by number and weighted percentage of values missing: 2018 101
Table 20. PETS-SR data products: 2018 111
Table 21. Reserve code values: 2018 113
Figure 1. Longitudinal design for the NCES secondary longitudinal studies program: 1972-2025 12
Figure 2. Longitudinal design for the HSLS:09 9th-grade cohort: 2009-2025 14
Figure 3. Keying and Coding System Degrees page: 2018 34
Figure 4. Keying and Coding System Courses page: 2018 34
Figure 5. Code diagram: 2018 36
Figure 6. PETS-SR unknown eligibility adjustment construction: 2018 60
Figure 7. PETS-SR nonresponse and calibration weighting adjustment construction: 2018 62
Figure 8. Example SAS-callable SUDAAN code to calculate an estimated mean and linearization standard error for a postsecondary transcript student-level analysis 80
Figure 9. Example SUDAAN code to calculate an estimated mean and replicate (BRR) standard error for a postsecondary transcript student-level analysis 80
Figure 10. Example Stata code to calculate an estimated mean and linearization standard error for a postsecondary transcript student-level analysis 81
Figure 11. Example Stata code to calculate an estimated mean and replicate (BRR) standard error for a postsecondary transcript student-level analysis 81
Figure 12. Example SAS code to calculate an estimated mean and linearization standard error for a postsecondary transcript student-level analysis 81
Figure 13. Example SAS code to calculate an estimated mean and replicate (BRR) standard error for a postsecondary transcript student-level analysis 81
Figure 14. Example R survey package code to calculate an estimated mean and linearization standard error for a postsecondary transcript student-level analysis 82
Figure 15. Example R survey package code to calculate an estimated mean and replicate (BRR) standard error for a postsecondary transcript student-level analysis 82
Figure 16. Example IBM SPSS complex samples code to calculate an estimated mean and linearization standard error for a postsecondary transcript student-level analysis 82