Brought to you by SchoolHouse Connection and Poverty Solutions at the University of Michigan.
The Data-to-Action Playbook: What It Is, and How to Use It
Section 1: Key Concepts and Lessons from the Field
- Chapter 1: Identify Your Goals and the Available Data
- Chapter 2: Vision
- Chapter 3: Leveraging the Data that is Available
- Chapter 4: Data Analysis & Data Presentation
- Chapter 5: Partnership & Team
- Chapter 6: Messaging
Section 2: Step-by-Step: How to Access and Use Data on Student Homelessness
- Chapter 7: Not Sure Where to Begin? Start with Publicly Available Data!
- Chapter 8: Some Analysis Required
- Chapter 9: Diving into Individual Level Student Data: High Rewards, Lots of Work
- Example Documents
- seeks to raise awareness of key challenges faced by children, youth, and families experiencing homelessness, and the systems that serve them; and
- aspires to increase implementation of practices and policies that improve educational outcomes, from early childhood through postsecondary.
Goals of the Campaign
- Early Childhood: By 2026, young children experiencing homelessness will participate in quality early childhood programs at the same rate as their housed peers.
- Graduation: A 90% high school graduation rate for students experiencing homelessness by 2030.
- Postsecondary: A 60% postsecondary attainment rate for students experiencing homelessness by 2034.
- Black high school students are 2.5 times more likely to experience homelessness than their white peers, Hispanic students are twice as likely, and American Indian/Alaskan Native students are 1.7 times as likely. 
- Students who identify as LGBQ are more than twice as likely, and transgender students are 9 times more likely, to experience homelessness. 
- High school students who become pregnant are 10 times more likely to experience homelessness, and students staying in shelters are 20 times more likely than their housed peers to become pregnant. 
Using these data to actively address inequities is imperative. As a first step, consider these questions:
- Does your district, institution, or state collect and disaggregate student homelessness data by race, ethnicity, sexual orientation and gender identity?
- If yes, what do those data show about the experience and outcomes of students who experience homelessness in your district or school?
- What additional data related to student homelessness and race, ethnicity, sexual orientation and gender identity do you need? For example, are high school graduation rates disaggregated by homelessness and race and ethnicity?
- How will you share this data, and with whom?
A Note about COVID-19 and Student Homelessness Data
Statewide Initiative - Building Changes (Washington)
- Why is this work important?
- Who are the individuals and groups I am trying to serve?
- What am I trying to accomplish through the data I am presenting?
Gather Appropriate Data
- Data analysis does not need to be complicated to have an impact. Often, it is the most basic information—like knowing the percentage of students who are homeless by school district—that has the largest impact.
- The more levels/geographies to which you can make the data relevant, the more people you will be able to engage/organize. People care about information that is local to them; this is particularly true when thinking about getting media attention.
- If you identify potentially useful data that you may not be able to obtain, consider alternative data sources that you may be able to use to convey your points. For example, if you can show the percentage of students who are homeless by district, but you do not have local proficiency rates for English or Math, you can point to national data on English and Math proficiency rates of students who are experiencing homelessness. Making this type of connection helps people understand the educational impact of homelessness and why it is an important local issue.
Statewide Initiative - Poverty Solutions (Michigan)
Poverty Solutions is a university-wide presidential initiative at the University of Michigan that partners with communities and policymakers to find new ways to prevent and alleviate poverty through action-based research. Their data analysis and research about students experiencing homelessness addresses identification, graduation and dropout rates, early childhood homelessness, and school discipline.
- Starting with public data can help you build a case that the issues you are raising merit closer attention.
- By identifying gaps in the data, you may be able to attract funding for further research.
- When possible, use indicators and methods that have been proven to be reliable.
- This will enable you to connect your findings to other work and build on what has already been demonstrated elsewhere, rather than having to start from scratch.
- Ask yourself which questions can and cannot be answered with the available information.
- Who is not included in the data and why? Are there potential issues of under- or over-counting? How would these issues potentially impact your findings?
- This is also the time to ensure your approach is comprehensive. For example, when looking at long-term impact, be sure to look at metrics other than graduation rates, such as highest degree attained.
Statewide Initiative - Indiana Youth Institute & School on Wheels (Indiana)
- What data is easy for you to process based on your team’s skill set?
- What outcomes and comparison groups will be most useful in demonstrating your significant points?
- Whose assistance do you need to facilitate your analysis of the data?
- How accustomed is your team to storytelling?
Complement quantitative with qualitative data
Meaningful comparison groups are critical to the story
It is also often fruitful to compare data for students who are economically disadvantaged  to data for students who are experiencing homelessness. This type of comparison can help people understand that homelessness has an impact over and above that of poverty alone. When possible, it is also essential to include demographic information—in particular, information about race and ethnicity —because students of color are disproportionately impacted by homelessness.
Addressing the issue of underidentification
- What type of training takes place for the school personnel who identify students experiencing homelessness and what do the identification processes look like?
- What kind of work has or has not been going on to improve identification in your school district or state?
- What steps have been taken to address the reluctance of children, youth, and families experiencing homelessness to self-identify?
- No matter what training or processes are in place, it is important to acknowledge that school-based numbers are probably missing a lot of children and youth who are homeless.
In light of these considerations, be candid about the limitations of the data. Keep in mind, however, that it is possible to learn from imperfect data.
Data without comparison lacks direction
Data that is engaging often invites diverse interpretations and responses
- What does the data show? (What?)
- Why should someone care? (So what?) Be sure to think about what your intended audience cares about. What are their priorities? What comparison groups will most clearly show how and why the experience of homeless students matters in the context of the audience’s priorities. If you cannot come up with a compelling reason in your own mind for why someone should care about the data, it is not useful data.
- Given the data, what systems are not working the way they should? (Now what?) What actions can be taken to address the issue? What are your programmatic and/or policy recommendations? When deciding on your recommendations, be sure to consider whether the people to whom you are targeting the presentation can make the changes that you are advocating.
The more that you can target your data and presentation to your audience, the more engaging it will be and the more it will invite interpretation and action.
Statewide Initiative - Research for Action (Pennsylvania)
- Identify a local school district as a partner.
- If you are unable to communicate a message publicly, partner with a person or an organization that can.
Communication with your team members and with collaborators is crucial. Discuss progress, challenges, and successes, and create opportunities for communication to ensure that relationships are mutually beneficial. Finally, remember that self-care is different for everyone, and may look different for the different people with whom you are working. In order to continue supporting students, practitioners must also take care of themselves and their teammates.
Diversity in data skills and cross-sector knowledge maximize success
Look for potential partners already doing this work
Advocacy efforts don’t have to be siloed
Set up communication systems, reflection time, and feedback loops
Statewide Initiative - The Learning Policy Institute & UCLA Center for the Transformation of Schools (California)
- Articulate your goal(s) and what each entails.
- Identify who has the decision-making power relating to your goal.
- Look at the level of support you need and what you might need to do to reach supporters.
- Map out specific names of key supporters and subordinate decision-makers, and then figure out who influences those individuals and where those influencers stand on your issues. If possible, find individuals and organizations who will be easy to connect with—this can help you hone in on where to begin your partnerships.
- At what level does policy need to change? School, district, state, national – all of the above?
- What component(s) of the policy need(s) to change?
- How will the change benefit students?
First, identify what needs to change, and then identify the information that will persuade people to act. Then make sure to suggest concrete policy changes/actions to address the problem.
- Policymakers value testimonials and constituent perspectives, but in the context of quantitative information. What are the numbers, the outcomes, the trends?
- Policymakers need help understanding the meaning of the data. That’s where the testimonials come in—students and families are the best messengers to give life to these numbers.
- For policymakers and people in the business world, brevity is also key; otherwise, what you’re communicating may not even be consumed.
- In the philanthropy sector, individuals tend to respond to objectively measurable metrics and outcomes.
- Media tend to be responsive to both testimonials and metrics, but media focus principally on the story that you are trying to tell.
- Think about ways you can leverage social media and live streaming.
It can be very helpful to present interactive data, which provides the flexibility to make the data most relevant to your circumstances. The more you are able to customize and tailor, the better.
Local Initiative - Inclusive Economy Lab at the University of Chicago (Chicago, Illinois)
- “Experiences of homelessness in early childhood are associated with delays in children’s language, literacy, and social-emotional development.” 
- “Early childhood programs prevent the harmful life-long effects of homelessness on education, health and well-being.” 
- The city should support collaborations between local shelters and early childhood centers to both identify opportunities for shelters to become more child friendly and to encourage families to enroll in early childhood programs. This type of approach has been successfully implemented in Philadelphia, PA. 
Real World Example
- Using the US Department of Education report “Early Childhood Homelessness: 50 State Profile”, which states are most effectively reaching young homeless children with their early childhood programs? What are the similarities and differences between your state and the states with more effective outreach? Are there people with whom you can speak in those states to find out what they are doing that is working? With whom in your state could you share this information?
- If you find that more granular detail is needed to engage people in your state on the issue of early childhood homelessness, you can combine this statewide data with your state’s K-12 data by school district (which we will discuss next). This type of analysis enables you to make policy recommendations such as:
- “An estimated 8% of young children in Oklahoma were homeless in 2018. Based on K-12 data on student homelessness, school districts with likely issues of under-identification for K-12 students are (list the districts). Exploring outreach efforts in these districts is a logical first step to improving early childhood enrollment for young homeless children in the state.”
Local Initiative - Camden City School District (Camden, New Jersey)
- “Kentucky has a higher rate of student homelessness than the national average (3.6% vs. 2.3% respectively), but only 8% of LEAs in the State are supported by McKinney-Vento Subgrants compared to 23% of LEAs nationally.”
- “Only 51% of homeless high school students in North Dakota graduated in four years, compared with 84% of homeless students in Kentucky.”
- “In the State of Michigan more than 4,700 youth are facing homelessness on their own, unaccompanied by a parent or guardian. These youth face greater risk to their health and well-being than their housed peers; they experience seven times the rates of suicide and five times the rate of sexual dating violence.”
- “School districts without McKinney-Vento subgrants may have greater challenges identifying students experiencing homelessness.”
- “Better graduation rates for high school students who are experiencing homelessness are possible.”
- “Michigan is among the minority of states that do not allow most unaccompanied minors to consent for their own basic health care.”
- “Given the overall higher rate of homelessness in the State and the lower proportion of LEAs with McKinney-Vento subgrants, the State of Kentucky should look at how rates of identification vary between school districts with and without McKinney-Vento subgrants.”
- “Outreach should be conducted to Kentucky and other states with higher graduation rates among students experiencing homelessness to identify practices that are helping students succeed.”
- “Amending Michigan’s state laws to permit unaccompanied homeless youth to consent to their own health care would keep them safer and may empower them to access education and employment that can end their homelessness.”
Real World Example
Other Simple Data to Explore for K-12 Homelessness:
Data and information relating to a broad range of homeless educational issues are analyzed and summarized by NCHE. The latest NCHE report summarizing such data can be found at Student Homelessness in America: School Years 2017-18 to 2019-20. NCHE also maintains K-12 homeless education data going back to the 2006 school year on its data summary page.
The NCHE Federal Data Summary reflects state-specific data that has been deduplicated. EDFacts Repository. Local educational agency data can be found in EDFacts data files at EDFacts Data Files.
Here are some questions to explore using your state’s data and the information in the NCHE Federal Data Summary (complemented in some cases by information that can be derived from the flat files available at the EDFacts website or other readily available data).
- What are the grade level proficiency rates in English and Math for students experiencing homelessness in your state compared to other states? Compared to all students in your state? Data for all students can be found on your state’s Department of Education website. The most recent information for homeless student proficiency can be found in the National Center for Homeless Education Federal Data Summary: School Years 2016-17 through 2018-19
- How many students are homeless in your state? Has the number increased or decreased in the last three years? Is there an opportunity to use these numbers to improve access to health or other services by connecting these numbers to health outcome data for homeless students? The most recent information can also be found in the Student Homelessness in America: School Years 2017-18 to 2019-20 and in SchoolHouse Connection’s report on the health risks of homeless youth.
- What percentage of enrolled students are homeless in your state? This information can be found in the National Center for Education Statistics’ Annual Reports’ Digest of Education Statistics. Relevant data also is available from NCHE.
- What proportion of homeless children in your state are living in shelter, doubled up, unsheltered, and in hotel/motels? Student Homelessness in America: School Years 2017-18 to 2019-20 provides relevant data.
- Has the number of unaccompanied youth in your state increased or decreased in the last three years? This information can also be found in Student Homelessness in America: School Years 2017-18 to 2019-20.
If you are not sure what questions you are interested in exploring with the data you have, below are reports that can be useful in formulating “So What” statements for helping stakeholders understand why it is important to meet the needs of students who are experiencing homelessness. When you identify an issue that you think will resonate with your local stakeholders, circle back to your data to see if there is evidence for further exploration around that issue.
- The Youth Behavioral Risk Factor Survey (YRBS). Currently, many states include questions on homelessness as a part of this survey and in 2021 all states will include homelessness as an risk indicator. National analysis of the health and behavioral risks faced by homeless students can be found at SchoolHouse Connection’s website in its report on lessons learned from the YRBS.
- Chapin Hall’s website has a large number of reports on unaccompanied youth homelessness and is an outstanding resource for anyone interested in addressing homelessness among unaccompanied youth in their state.
- Hidden in Plain Sight is another report that provides context and compelling details on the experiences of youth who face homelessness in America’s public schools.
Did You Know: The Every Student Succeeds Act (ESSA) requires that Local Educational Agencies (LEAs) report graduation and dropout data separately for students who are experiencing homelessness. As of December 2021, graduation and dropout data for homeless students in your state can be found in the National Center for Homeless Education Federal Data Summary: School Years 2016-17 through 2018-19. NCHE will release a separate report with updated graduation statistics in 2022.
Local Initiative - The People’s Emergency Center (Philadelphia, Pennsylvania)
FAFSA and Homeless Youth: Challenges and Recommendations in the COVID Era
Real World Examples
- Reading High School (Pennsylvania) hired a success coach to follow its students from high school into college and to facilitate the students’ post-secondary educational success, including through forging partnerships between the high school and local colleges and universities to share data and information about students experiencing homelessness. You can learn more in this SchoolHouse Connection archived webinar.
- In 2019, a RealCollege survey found that 17% of the responding students had experienced homelessness during the previous year. The survey reported data from 167,000 respondents from 171 two-year and 56 four-year institutions.
- The Hope Center conducted a similar survey during the pandemic and found that 11% of students at two-year institutions and almost 15% at four-year institutions had experienced homelessness due to the pandemic. This was taken from a sample of over 38,000 students attending 54 colleges and universities.
Local Initiative - Black Male Institute at UCLA (Los Angeles, California)
- Which ten districts in your state have the highest numbers of students experiencing homelessness? Does this list make sense based on the total student enrollment and the total number of children living in poverty in each district?
- Which ten districts in the state have the highest percentages of students experiencing homelessness? Are these districts the same or different from the districts with the largest number of homeless students?
- Which districts have the highest numbers of children eligible for free or reduced-price lunch? Are these districts the same as the districts with the highest numbers of homeless students? If they are not the same, is there a logical reason why they may differ, or is this evidence of a likely undercount?
- Are there districts with high rates of poverty but low rates of homelessness? If so, is there a logical reason for this, or is it worth exploring whether there has been an undercount of students experiencing homelessness?
- What does a map of the percentages of students experiencing homelessness by school districts in your state look like?
- Is the percentage of low-income students who are experiencing homelessness similar across the state, or does it vary widely? If it varies, does this variation make sense based on what you know about the affected districts?
- Which ten districts have the highest numbers of unaccompanied homeless youth?
- What are the staff-to-student ratios in districts with a higher percentage of students experiencing homelessness compared to districts with a lower percentage of such students?
This data set also allows you to look at the percentage of homeless students by primary nighttime residence, disability status, English language learning needs, and other school district indicators. The local nature of the data facilitates targeted outreach, and also can be useful for communicating with local media. If your organization has access to someone with mapping skills, another benefit of this data is that it is fairly easy to create a map that will enable visual comparisons across the state. An example of a map using this data can be seen below.
- Go to the data set.
- Select Download Data.
- Open the Excel file and use the filter function to select only your state.
- Copy the selected data and paste it into a new Excel tab.
- Once the data is in a new tab, select all of the data in the sheet, go to the data function at the top of the excel sheet and select “Sort.”
- Start by sorting by the number of homeless students.
- Then sort by the percent of homeless students.
- Then sort by the number and percent of students who are eligible for free or reduced-price lunch.
- As you are sorting the data, take notes.
- What patterns do you see, what makes sense, and what does not make sense? Often, what does not make sense only becomes obvious when you undertake comparisons.
- Remember no data is perfect, sometimes by looking through the data you may identify a place where the numbers have been recorded inaccurately, or where there is a large undercount of homeless students taking place. Bringing these to the attention of your community can be important in moving forward a conversation.
- You can also sort by the number of unaccompanied youth and the percent of students with disabilities who are identified as homeless.
- From your notes, write down “What” statements. For example:
- The ten districts with the largest number of homeless students were….
- While ____ has the largest number of students receiving free and reduced-price lunch, it has a smaller number of homeless students than _____, which has fewer students receiving free or reduced-price lunch.
- ___school district identified less than 3 homeless students as having special education needs.
- __ identified only 11 unaccompanied youth.
- To decide which issue(s) to focus on, follow these “What” statements by writing down why each of these facts is significant for your community. For example:
- How does this data compare to other points of information and what does that suggest?
- Where does the data point to there being the greatest need for services?
Real World Example
- Go to your state Department of Education website.
- Look for a tab or check the website’s index for references to “data,” “analytics,” “reporting,” “accountability,” or something similar.
- From here, some states will have an interactive tool for you to use, while other states will have links to different Excel files. If there is an interactive figure generator, you can be relatively confident that if you search around enough, you can usually find an underlying Excel table organized by school district. Finding this may require some time.
- The best indicator to explore first is Graduation and Dropout rates. This is because under the Every Student Succeeds Act (ESSA), all states are required to list graduation and dropout rates for homeless students. If there is any publicly available data from your state’s Department of Education that is broken out by housing status, it will, at a minimum, include graduation and dropout rates.
- The next two indicators that you most likely will be able to find are English and Math proficiency rates and chronic absenteeism. Again, not all states will have these available, but these indicators are the most likely to be included.
- If you do not find these indicators in a publicly available format on your state’s Department of Education website, the next step is to send an email request to the person in charge of data for the state. Your request should be specific. Start by asking for Graduation and Dropout rates, ELA and Math Proficiency rates, and Chronic Absenteeism rates for: homeless vs. non-homeless students, economically disadvantaged/free or reduced-price lunch eligible students, and non-economically disadvantaged/free or reduced price lunch eligible students. Ask for these indicators broken out by these groups at the School District and Intermediate School District (ISD) level. Requesting both geographical breakouts will provide you with more flexibility in case the school district geographic level has a lot of redacted data (data removed to protect the privacy of students due to small numbers).
- Examples of what this process looks like in a few different states are set forth below:
- The Michigan Department of Education (MDE) houses its data at https://www.mischooldata.org/
- To find attendance rates for students experiencing homelessness in kindergarten through high school, you would do the following:
- Click K-12 tab
- Report Category (Homeless)
- The Idaho Department of Education houses its data on its website at https://www.sde.idaho.gov/
- To find graduation rates for students experiencing homelessness in kindergarten through high school, you would do the following:
- Click Menu
- Navigate to Departments > Assessment & Accountability
- Assessment & Graduation Rate Results
- Scroll down to Graduation Rates and click on it
- Click “Four Year Graduation Rate Master” to download file
- Once you open the file, you will be able to see the graduation rate of different student populations.
- The California Department of Education houses its data at https://www.cde.ca.gov/ds/ (student-level data is available within DataQuest)
- DataQuest affords flexibility in pulling a data set. For example, it provides the option to filter by subgroup: socioeconomically disadvantaged, migrant, foster, or homeless. Viewing it may provide ideas for advocacy in your own state, including advocacy urging education departments to make various data sets readily available.
- The micro, student-level allows for deeper analysis than meso- and macro-level data, because it provides flexibility to explore different trends, such as:
- Homeless prevalence data by geographic areas
- Personalized outcomes data.
- Examining this data also allows for more tailored approaches to identifying student needs.
- Further, it can be used to track the long-term impacts of homelessness on students over an extended period of time, rather than at a single point in time.
- In turn, demonstrating these trends can help focus conversations on what students are actually experiencing.
- A data-sharing agreement with your state will be required to obtain the data.
- Some states have instructions and forms on their website for obtaining the data, while other states do not.
- If you are interested in using individual-level data, you will likely need a research partner who has the capacity to analyze and store the data securely.
- Research institutions may have the ability to house data in a way that would be FERPA-compliant. Be aware of any processes you might need to follow, such as filing paperwork for the Institutional Research Board (IRB).
- Percent and Number of Homeless Students by School District (all states)
- Chapin Hall Data
- Data Dashboard (Building Changes Washington)
- Processed homelessness data by state
- National Association for the Education of Young Children (NAEYC)
- National Center for Education Statistics (NCES)
- National Center for Homeless Education (NCHE)
- Homelessness Data Exchange Website
- Health Outcomes of Homeless High School Students
- Youth Behavioral Risk Factor Survey Raw Data by State (CDC) 
- Course access data, enrollment, classes
- Other state systems (Medicaid, corrections, etc.)
- Postsecondary data
- Workforce and career data
- Jennifer Erb-Downward, Senior Research Associate, University of Michigan Poverty Solutions (Michigan)
- Matt Lemon, Interim Director of Research & Evaluation, Building Changes (Washington)
- Ian Rosenblum, former Executive Director, The Education Trust – New York
- Daniel Zavala, Interim Executive Director / former Director of Policy & Strategic Communications, Building Changes (Washington)
 Student Homelessness: Lessons from the Youth Risk Behavior Survey Part III: Sexual Orientation and Gender Identity Inequity (2021, SchoolHouse Connection)
 Student Homelessness: Lessons from the Youth Risk Behavior Survey Part X: Pregnancy Rates of High School Students Experiencing Homelessness (2021, SchoolHouse Connection)
 Economically Disadvantaged Students are those eligible for free- or reduced-price meals under the National School Lunch program, are in households receiving food (SNAP) or cash (TANF) assistance, are eligible under Medicaid, are homeless, are migrant, or are in foster care.
 Ziol-Guest, K.M. & McKenna, C.C. (2014) “Early childhood housing instability and school readiness,” Child Development, 85(1), 101-113.
 See https://www.pec-cares.org/building-early-links-for-learning.html
 Note: if your state does not yet include questions on homelessness, these questions can be added to the next round. Reach out to SchoolHouse Connection for more information.