Appendix

Scope and Methodology

The Joint Legislative Audit Committee (Audit Committee) directed the California State Auditor to examine LEAs’ efforts to identify, provide services to, and support students from preschool and kindergarten through grade 12 who are experiencing homelessness. Specifically, the Audit Committee directed us to determine the roles and responsibilities of the State Board of Education and Education in overseeing the McKinney‑Vento Act, determine the process that selected LEAs use to identify youth experiencing homelessness, and identify best practices for identifying these youth. The table below lists the objectives that the Audit Committee approved and the methods we used to address them.

Table
Audit Objectives and the Methods Used to Address Them
AUDIT OBJECTIVE		METHOD
1	Review and evaluate the laws, rules, and regulations significant to the audit objectives.	Reviewed relevant federal and state laws, rules, and regulations related to identifying and supporting youth experiencing homelessness.
2	For Education and the State Board of Education, determine the following:
	a. Their roles and responsibilities in administering the McKinney‑Vento Act.	Interviewed officials at Education and the State Board of Education and obtained documentation to understand their respective roles and responsibilities Reviewed appropriate documentation to determine whether the State Board of Education and Education complied with their responsibilities under the McKinney‑Vento Act.
	b. The guidance, if any, they provide to local educational agencies concerning the McKinney‑Vento Act.	Interviewed staff of the state coordinator and the State Board of Education and reviewed available documentation that Education communicates to LEAs and has posted on its website. Reviewed documentation from trainings Education conducted for LEAs from academic years 2015–16 through 2017–18. Assessed the adequacy of the guidance and training that Education provides to LEAs.
	c. The steps these entities take to give support to and ensure compliance by LEAs that report no youth experiencing homelessness among their students.	Interviewed Education and State Board of Education staff and reviewed available documentation to identify Education’s methods for monitoring LEA compliance with homeless education requirements in state and federal law. Selected and evaluated federal program monitoring reviews for six LEAs that Education conducted during fiscal years 2015–16 through 2017–18 to determine whether Education complied with its established monitoring processes. Assessed additional steps Education has taken to monitor and provide technical assistance to LEAs that report zero youth experiencing homelessness.
	d. The steps, if any, they take to collaborate with other state, local, and national entities that deal with youth experiencing homelessness. Determine whether barriers exist to this collaboration and recommend solutions for how to remove them.	Interviewed staff of the state coordinator and the State Board of Education to identify the methods and frequency with which Education collaborates with other entities that serve youth experiencing homelessness and to identify any barriers it faces. Reviewed documentation from those collaborations to determine the purposes and effects of each collaborative effort. We found that Education collaborates with various entities as required and we did not identify anything that suggests any barriers to collaboration.
3	Identify for the past three fiscal years the number of youth experiencing homelessness identified at each of California’s LEAs. Evaluate these data and determine whether factors such as student population, geographic area (including urban, suburban, and rural areas), or other demographic factors affect LEAs’ abilities to identify the number of youth experiencing homelessness.	Obtained CALPADS data available from Education on the number of youth experiencing homelessness that LEAs identified for academic years 2015–16 through 2017–18. Analyzed the number of youth experiencing homelessness that each LEA identified from academic years 2015–16 through 2017–18 to determine trends in identifying these youth by LEAs in urban, suburban, and rural areas. We used these same data to evaluate the effects of student enrollment size and liaison turnover on the LEAs’ abilities to identify youth experiencing homelessness.
4	For one charter school and a selection of three to five LEAs—one located in each rural, urban, and suburban area, and at least one located in the San Francisco Bay Area and one located in San Bernardino County—that reported recently having zero or very few identified youth experiencing homelessness, and one LEA that has been successful in identifying and serving a high number of youth experiencing homelessness, determine and evaluate the following:	Using CALPADS data available from Education's website on the number of youth experiencing homelessness that LEAs identified for academic year 2017–18, we selected the following LEAs for review: Birmingham Charter, Greenfield, Gridley, Norwalk‑La Mirada, San Bernardino, and Vallejo.
	a. The methods the LEAs and charter school use to identify youth experiencing homelessness.	Interviewed staff and reviewed available documents at each selected LEA to understand how the LEA identifies youth experiencing homelessness. Reviewed documents to determine if the LEA uses methods that best practices recommend to identify youth experiencing homelessness.
	b. The activities McKinney‑Vento liaisons take to support the LEAs and charter school in identifying youth experiencing homelessness.	Reviewed the websites of each of the six LEAs we selected, as well as the websites for two of each LEA’s schools, to determine if the LEAs have provided sufficient and easily found information on the McKinney‑Vento Act, support for youth experiencing homelessness, and the contact information for the local liaisons for youth experiencing homelessness. Reviewed documentation from each LEA to determine if the LEAs provide translated versions of informational materials about the McKinney‑Vento program in required languages. Interviewed staff and reviewed available documents at each LEA to determine the frequency and content of McKinney‑Vento trainings.
	c. The support and technical assistance the LEAs and charter school receive from county offices of education and state entities.	Interviewed staff and reviewed available documents at each LEA to determine the support and technical assistance they receive from Education and their county offices of education. Interviewed staff and reviewed available documents at the six county offices of education that support the LEAs we reviewed to determine the support and technical assistance they receive from Education and provide to the LEAs in their respective counties.
	d. How the LEAs and charter school collaborate with other homeless service providers in identifying youth experiencing homelessness.	Interviewed staff and reviewed available documents at each LEA to determine if they collaborate with other service providers to identify youth experiencing homelessness.
	e. The barriers that exist to sharing data with other homeless service providers and possible solutions to these barriers.	Interviewed staff and reviewed available documents at each LEA to identify barriers that exist to sharing data with other homeless service providers and to determine what actions the LEAs have taken to address any potential barriers.
	f. Whether the LEAs and charter school apply for and use federal grants to help with their effort to identify and serve youth experiencing homelessness.	Reviewed documentation from Education and the LEAs to determine if the LEAs have applied for, received, and used the McKinney‑Vento subgrant. The two LEAs we reviewed that received grant funding generally used the funds for staff positions and services related to their homeless education programs.
5	To the extent possible, using information identified at the LEAs and the charter school, identify the following for LEAs: a. Trends and best practices in methods LEAs use to identify youth experiencing homelessness. b. Trends and best practices in activities that McKinney‑Vento liaisons take to support LEAs’ efforts to identify youth experiencing homelessness. c. Trends and best practices in support and technical assistance LEAs receive from county offices of education and state entities to identify youth experiencing homelessness. d. Trends and best practices in how LEAs collaborate with other homeless service providers to identify youth experiencing homelessness. e. Common barriers that exist to sharing data between LEAs and other homeless service providers and possible statewide solutions to these barriers.	Compared the methods, trainings, activities, support, collaboration, and barriers that exist at the six LEAs we reviewed to determine trends and best practices for identifying and supporting youth experiencing homelessness.
6	To the extent possible, identify national best practices for identifying and supporting youth experiencing homelessness.	Interviewed the homeless education coordinators at U.S. ED to determine best practices for identifying and supporting youth experiencing homelessness and to request recommendations for which states and nonprofit organizations to review for national best practices. Based on U.S. ED’s recommendation, we judgmentally selected five states—Florida, Texas, Georgia, Michigan, and Washington—and we judgmentally selected three nonprofit organizations that work with youth experiencing homelessness to interview.
7	Review and assess any other issues that are significant to the audit.	Using Education’s online DataQuest tool, at a statewide level and at the LEAs we selected for audit objective 4, compared the rates of suspension, chronic absenteeism, dropout, and graduation during academic year 2017–18 for youth experiencing homelessness with those of their peers to determine whether the homeless education program is effective. To determine if the LEAs provided the needed services to youth experiencing homelessness, we reviewed the LEAs’ budgets and expenditures and whether the LEAs tracked the services that each individual needed and provided those services. Nothing came to our attention to suggest that youth experiencing homelessness were not provided the services they needed.

Source: Analysis of the Audit Committee’s audit request number 2019‑104, state law, and information and documentation identified in the column titled Method.

Assessment of Data Reliability

The U.S. Government Accountability Office, whose standards we are statutorily obligated to follow, requires us to assess the sufficiency and appropriateness of computer‑processed information we use to support our findings, conclusions, or recommendations. In performing this audit, we relied on reports obtained from CARS and CALPADS.

We did not conduct a comprehensive data reliability analysis of the reports from the CARS and CALPADS systems because the supporting documentation is maintained among California’s approximately 2,300 LEAs, making accuracy and completeness testing impractical. To gain reasonable assurance of the data’s completeness and accuracy to support our audit findings, we reviewed other information provided by Education. Specifically, to determine if the CARS data were complete, we compared the number of LEAs listed in CARS to the number of LEAs listed in CALPADS. Further, to determine the accuracy of the data contained in CARS, we used Education’s Title I report for academic year 2017–18 to identify the Title I funds each LEA received and compared it to the amounts the LEAs reported receiving for the same year in CARS. We determined that the CARS data were unreliable for academic year 2017–18. Further, to gain reasonable assurance of the completeness of the data in a report that Education generated from CALPADS for cumulative end‑of‑academic year student data that it does not publish on its website, we compared the number of LEAs in Education’s CALPADS point‑in‑time data that it publishes online to the number of LEAs within the report. We found that Education reported 18 more LEAs in the cumulative end‑of‑year data than in the point‑in‑time data it reports on its website. Therefore, we concluded that the CALPADS data that Education provided us were of undetermined reliability. Although we recognize that any limitations that we identified in the CARS or CALPADS data may affect the precision of the numbers we present, there is sufficient evidence in total to support our audit findings, conclusions, and recommendations.