Exploring Trends in Your Local YRBS or Student Health Survey Data

By Lisa N. Arsenault, PhD and Stefanie Albert, MPH

 

The CDC’s Youth Risk Behavior Surveillance System (YRBS) has provided data on health-related behaviors of U.S. high school students for over 25 years.  Results have been used to monitor progress toward national health objectives and to support the modification or development of programs and policies that promote health among high school aged youth.

Individual school districts across the country have recognized the utility and power this type of data provides and many choose to implement local-level versions of the YRBS (or other similar type of student health survey).  Most districts hire external contractors or organizations to conduct the data cleaning and analyses of their local data. Many districts also use external contractors to generate summary reports, tables, and charts.  But even with external help it can be overwhelming to decide how best to examine the results.

 Here we briefly illustrate one of the most powerful ways to explore student health survey data, the trend over time. Our goal is to provide some useful suggestions for exploring and reporting data from your own surveys so that you can be better informed on the health needs of your students. 

 Example 1:  The Basic Trend

Here is a simple table of results that shows the rate of one risk behavior among all students over three different years.  In this format it is easy to see and report if the rate of the behavior is increasing or decreasing over time. In our example, the rate of marijuana use in the prior 30 days has declined from nearly 27% in 2012 to about 19% in 2016.   

You can also visualize this data in a very simple bar chart, as shown below. You can opt to add some text boxes that highlight extra information that might be important for your audience to understand about the data such as the state rate, total number of students surveyed, etc. 

In this example, we added the approximate number of students the current year’s rate represents.  We find this is often helpful when trend data show a decrease over time because it’s easy to forget you are talking about actual students in your school.  So highlighting how many students are still at risk is a good way to balance the ‘big picture’ with the real personal value of the data. 

 Examining the overall trend in your data is absolutely the essential first step.  However, the trend you see might not represent what is going on for all sub-populations of students!  For this reason, we highly recommend exploring and comparing the trends among sub-groups of students to gain a more accurate picture of the health of your student population. 

 

Example 2:  Trend by Sub-Group

Here is a second table of results that shows the rate of one risk behavior over three different years and stratified by grade level. In this format it is easy to see and report if the rate of the behavior is increasing or decreasing for each sub-group.  In our example, the rate of marijuana use in the prior 30 days has steadily declined between 2012 and 2016 for 9^th and 12^th grade students.  But the rate has increased since 2012 among 11^th graders and it has increased since 2014 for 10^th grade students.  Had we stopped exploring our data after looking at the overall trend, we would have missed this very important finding!

 

Visualizing the trends by sub-group can really help you ‘see’ the differences, particularly when many years of data make reading summary tables full of numbers more difficult. In our example below, the dramatic drop in 30-Day marijuana use among 9^th graders is very clear.  Likewise, you can easily see the steady decline in use among 12^th graders.  And the less encouraging results for the 10^th and 11^th grade students are shown in a more understandable way that can foster discussion with stakeholders or audiences. 

 

 

 Trend data is one of the most powerful ways to explore your student health survey data.  But overall trends may be hiding some very important differences between sub-groups of students.  At ICH we always encourage school districts to look at data trends by grade level and by gender.  Additionally, for districts with a diverse student population looking at data by race/ethnicity is also important.  Ultimately, the goal of collecting YRBS or student health survey data is to inform programs and policies that will improve the health and wellbeing of all students.  We hope that our suggestions here will help you achieve that goal.   

 

*********************

 ICH has supported the development and implementation of both middle school and high school student health surveys in multiple school districts in MA. We have also provided technical assistance to numerous school districts, public health agencies, and substance abuse coalitions around the analysis, interpretation, and dissemination of survey results.  Our goal is always to aim for results that are understandable, useful, and actionable and this has led us to explore many different ways to visualize student health survey results over the years.

 

Learning About My Community Through Data: Reflections on My Summer at ICH

By Grace Chan

Tufts University MS in Nutrition/MPH Candidate

My summer as the Data Analyst Intern at the Institute of Community Health (ICH) has been an enriching experience. ICH has provided me insight into the world of quantitative data as well as taught me more about the communities surrounding metro-Boston – both aspects I wanted in a summer internship.

Before joining ICH in May, I finished my first year of graduate school at Tufts University.  As a MS Nutrition/MPH student, my goal is to gain more knowledge about the role of nutrition interventions in preventing chronic diseases. I want to gain skills in epidemiology and biostatistics in order to understand the complex interactions between nutrition, environmental exposures, and health behaviors that lead to various health outcomes. Having lived in Boston for less than a year, I decided to stay here to complete my summer internship rather than go home to California, in order to learn more about the community I am now a part of.

I first heard about ICH through a course at Tufts.  I was intrigued by the work ICH does in collaboration with local communities, and was excited to see a data analyst internship position available for the summer.  At ICH, I have had the pleasure of working on the Everett Data Book Project with Lisa Arsenault, Blessing Dube, Kelly Washburn, and Reann Gibson.  We worked with the Everett community representatives, Jean Granick, Bob Marra, and Jamie Stein, to compile a health assessment data book, detailing the health of the community’s adult and youth populations. Once completed, the data book can be used to inform the development of community programs and promote policy change.

During the internship, I learned to access data from various sources, such as MassCHIP, the Massachusetts Department of Education, and the Bureau of Labor Statistics.  Furthermore, I learned to analyze surveys such as the Youth Risk Behavior Survey (YRBS) and the Behavioral Risk Factor Surveillance System (BRFSS).  I gathered data of interest to the community members from these sources, organized and cleaned the data, and created charts that display the information in usable and meaningful ways.  Additionally, I participated in meetings with the Everett community representatives and made further improvements to the charts based on their feedback.  My team gave me a lot of support and guidance, plus the independence to explore the data and create charts from information I believed the community would find important.  They also offered me opportunities to present the data, most notably, to the larger work group in Everett that represents various sectors in the community.

As I reflect on my time at ICH, I know that I’ve gained valuable knowledge and skills that I will use in my future endeavors.  For example, I am more comfortable working with (and less intimidated by) large data sets.  Also, I developed a keener eye for detail and gained lots of experience working in Microsoft Excel.  Most importantly, I experienced collaborating with community members whose insights give more meaning to the data.  Overall, I am glad to have spent my summer with the wonderful members of my team.  I enjoyed getting to know my ICH team members and the Everett community.  It has been rewarding to be able to contribute to the process of creating the data book, and I cannot wait to see the final product!

Learn about the ICH Internship Program

  -----------------------------------------------------------------------------------------------------------

The views expressed on the Institute for Community Health blog page are solely those of the blog post author(s), and do not necessarily reflect the views of ICH, the author’s employer or other organizations with which the author is associated.

Using the Census to Make Sense of Your Community

    By Lisa Arsenault, PhD, Shalini Tendulkar, ScD, ScM, Nazmim Bhuiya, MPH, Kelly Washburn, MPH, Lise Fried, DSc, MS

The first population census of the United States was conducted back in 1790 when the population was just under 4 million.  The US Census Bureau has come a long way since those early days, yet the images that come to mind are still probably pretty archaic — data enumerators going from door to door with pencils and clip boards and maybe even on horseback in the early years!  Well today the Bureau works like a well-oiled machine, coordinating a monthly data collection of 250,000 households via mailed questionnaires, computer assisted telephone interviews, and in-person interviews.  This data (called the American Community Survey or ACS) is combined to give us very detailed estimates each year of who is living in the US.  In contrast, the decennial census (conducted every 10 years) collects only basic information on gender, race/ethnicity, household composition, and housing tenure.    

Recent news that the US House of Representative voted to cut funding for the ACS (and the Economic Census) may lead you to conclude that the data from such surveys cannot be that informative or useful.  However, nothing can be further from the truth, particularly for those of us working with communities and public health-related programs at the local level!  US Census Bureau Director, Robert M. Groves, does a wonderful job explaining the impact of the proposed defunding in his Director’s Blog.  Here we hope to highlight several examples of how the Institute for Community Health has used data from the American Community Survey to support the work of our community partners and further research into the health of local populations.

One nearly universal topic of interest is poverty.  The ACS is one of the primary sources for poverty data and we work on this topic frequently.  Recently, ICH pulled together a presentation for the Community Affairs Department of Cambridge Health Alliance that included the proportion of individuals living below the poverty line in Cambridge, Chelsea, Everett, Revere, and Somerville, MA.  We paired the data with information that ICH had collected on the weight status of adults in these cities and were able to illustrate a relationship between poverty and obesity at the population level.  The data provided a catalyst for the Community Affairs group to begin discussions related to food justice within their service communities.
Census data is also an integral part of our work as evaluators of public health programs.  The Massachusetts Alliance on Teen Pregnancy’s Youth First Initiative is currently testing community-wide approaches to reducing teen pregnancy in Springfield and Holyoke, MA.

As evaluators of their efforts, ICH has pulled data from the ACS to determine the number and proportion of teens residing in each city by census tract.  This data will be mapped and overlaid with the locations of the community health centers that have partnered with MATP on the project.  This information will provide Youth First with invaluable information on where the at-risk population resides within the community and how well the clinical providers are geographically positioned to serve the target population.   Such information can help a program evaluate and target their efforts, use resources more efficiently, and achieve a greater impact on a community’s health.

Finally, we frequently use census data to more generally describe a community’s population and its changes over time.  ICH supported the data efforts of the “Well Being of Somerville Report 2011” which was released last fall by the Somerville Health Agenda at Cambridge Health Alliance.  Using census data, we were able to describe the current population of Somerville, MA and show how the population’s characteristics have changed over the past decade.  Indicators including age, race/ethnicity, poverty, housing, language, country of origin, income, and educational attainment were all obtained from the US Census and ACS for Somerville.  And importantly, these indicators are all considered ‘determinants of health’ or factors in peoples lives that can affect one’s health in positive and/or negative ways.  Collecting these types of data and examining them alongside other types or sources of data yield great insight into the assets a community possesses as well as the challenges it may face from a public health perspective.  Overall, the this report is currently serving as a tool for local leaders and stakeholders to determine the public health needs of Somerville and inspire future planning efforts in the city.

We hope you’ve been impressed by how integral the data from the US Census Bureau is to what we do to improve health.  Our community partners rely on these data to inform their efforts — whether spurring new conversations, planning and targeting programmatic activities, or generally assessing the characteristics of a community.  There simply is no substitute for the American Community Survey.   And the most amazing part?   The data are there, right now, available online to anyone who is interested in learning about their own community.  Check it out!

______________________________________________________
The views expressed on the Institute for Community Health blog page are solely those of the blog post author(s), and do not necessarily reflect the views of ICH, the author’s employer or other organizations with which the author is associated.