Notes & Limitations
This page will receive additional updates in the coming months. If you have a question about the website or are confused about how the data is presented, please contact us at email@example.com. We plan to add additional questions and answers to this page based on user feedback.
Why are some states missing data?
Availability of state-level data was a key criterion guiding selection of the 21 Critical Health Objectives. While these objectives mark progress in this regard, challenges remain. All states collect vital statistics that cover the four mortality objectives (overall mortality, motor vehicle fatality, homicide and suicide). In this database, there are three main reasons why states may lack data on the other 17 objectives:
1. State-level data are not available. We were unable to locate state data for seven of the 21 objectives: alcohol-related crashes, children/adolescents with disabilities who are sad/unhappy/depressed, treatment for mental health problems, pregnancy, HIV/AIDS, Chlamydia, and overweight/obese. In some cases, similar data are available, but the available measure did not correspond to the official Healthy People 2010 measure. Please see Data Sources below for more detail.
2. States did not collect the data. Data collection varies tremendously for the ten behavioral objectives, monitored by the Youth Risk Behavior Surveillance System (YRBSS). Some states have final data only; others have baseline and final; and some collected data on a few objectives, but not all.
3. States use a data source other than YRBSS. Several states do not use the YRBSS. Many of these states use their own system to monitor trends in behaviors. Because of limited resources and because the measure may differ from the Healthy People 2010 objective, we do not present data from these other surveys. We provide links to these non-YRBSS surveys, where possible, and encourage users to visit those pages.
Why are data reported for only certain racial/ethnic categories?
Eliminating disparities, including racial/ethnic disparities, was one of two overarching goals for Healthy People 2010 and remains a major goal of Healthy People 2020. Therefore, this database places major emphasis on highlighting disparities, both in describing state data and making comparisons with national data. There are two major cases where discussion of different groups is limited:
1. State sample sizes are too small. In some states, the YRBSS sample size for some subgroups was too small to yield analyzable results. Similarly, mortality rates based on less than 20 deaths are considered unstable. This occurs most frequently in states with small non-White populations. Please see Data Presentation for more detail on how these issues are addressed in this database.
2. In state-national comparisons, we only compare racial/ethnic groups where state AND national data are available. In the Comparison with national data paragraphs, we do not comment on national groups with no corresponding state data. For example, Wisconsin has no stable mortality data for Asians; only for Blacks, Hispanics and Whites. While Asians have the lowest homicide rate nationally, we only compare the national rates for Blacks, Hispanics and Whites. A comprehensive summary of national data is available here.
Why aren’t there other disparities data?
This database only reports on gender and racial/ethnic disparities. It should also be noted that these data sources largely lack measures of socioeconomic status, such as education and income, which is strongly related to health behavior and health status for people of all ages, including adolescents. In addition, data are largely lacking for vulnerable populations of adolescents (for more information on vulnerable populations and data for the 21 Critical Health Objectives, please click here).
Why aren’t the most current data reported?
We selected national and state data to correspond with the baseline and midcourse timeframe established by Healthy People 2010. For most objectives, baseline is 1998-99 and final measures are from 2007-09. The Data Sources section provides detailed information about baseline and final years for the objectives and links to the most recent data. Since 1979, Healthy People has provided 10-year national objectives for promoting health and preventing disease. The 2010 objectives were established in the late 1990s. In 2010, the Healthy People 2020 objectives will be released along with guidance for achieving the new 10-year targets.
How does the summary text address missing data?
As described under Data Availability, each state has a unique combination of available data: some lack baseline data; some lack analyzable mortality or behavioral data for some racial/ethnic groups; and some collect data on selected objectives only. Here are a few highlights about how we treat missing/limited data:
Mortality data based on fewer than 20 deaths. Mortality rates based on less than 20 deaths are considered unstable. In cases where only one subgroup accounts for analyzable mortality, we focus our discussion on that group. (For example, if White 15-19 year olds are the only group in a state with analyzable data for suicide, the text described findings in terms of that group.)
No baseline data. When states have no baseline data, the highlights and comments focus on disparities and comparisons with national rates and disparities for the final year.
How do the data tables and the lists of objectives address missing data?
In short, if there are no data to analyze, the tables show an “X” (or “0.00” on some tables). For unstable mortality rates, we present that rate with an asterisk. An “X” in a table may indicate one of four situations:
1. Data weren’t collected.
2. There were not sufficient data to calculate a reliable rate or percentage. (Please see Data Sources for more detail)
3. For mortality, there were no deaths (eg., there were zero deaths from motor vehicle crashes for American Indian/Alaskan Natives ages 15-24 in Arkansas in 1999). (This may also be presented as “0.00”.)
4. There was not enough data to calculate a numeric change (for the third column of the tables).
On the list of objectives for each state, we only include active links to tables that contain data. If a link for an objective is “grayed out” (i.e., inactive), it means data are not available for that objective in that state.
Readers should bear in mind these points when comparing rates over time and between groups.
1. For behaviors, we present tables from YRBSS that show which changes over time are statistically significant. A rate may appear to increase or decrease over time, but statisticians also consider points, such as sample size, to determine if the change reflects true change in the population. The tables do NOT show which differences among gender or racial/ethnic groups are significant. The text emphasizes relatively large group differences and limits drawing conclusions about small differences. In this way, the text balances the importance of disparities, while recognizing that not all differences in the tables reflect true differences in the population.
2. While we recognize that rates based on few deaths are unstable, we also recognize that such small rates may represent progress in mortality. That is, mortality may improve (decrease) to the point where the rate is unstable. This makes it difficult for a state to claim progress, especially in small states when the number of deaths decreases substantially. In cases where the baseline rate was stable and the final rate is unstable, the text states that data suggest a decline.
Why is numeric change displayed and not percentage change?
We chose to present numeric, or absolute, measures of change. We chose not to present percentage, or relative, change because the percentage change of small numbers often over-emphasizes the scale of change.
What abbreviations are used for racial/ethnic classifications?
The summary text for each state and the national pages uses the following abbreviations:
White (“White, non-Hispanic” in most data sets)
Black (“Black, non-Hispanic” in most data sets)
Hispanic (specific to “Mexican” for national Overweight and Obesity data set, and “Latino” for national STD and Sex Behavior data set.)
Native American (“American Indian/Alaskan Native, non-Hispanic” in most data sets)
Asian/Pacific Islander (“Asian/Pacific Islander, non-Hispanic” in most data sets, and “Native Hawaiian or Other PI” for national STD data set)
What are the sources for National Data?
National data come from several sources. Please click here to look up the table listing national data sources. The data source for each objective is listed under the notes at the bottom of each page. For each objective, Healthy People 2010 determined the data sources, target, and year of baseline and midcourse measures. (Please see our Midcourse Review article for more information and references about national data sources and 2010 targets).
What are the sources for State Data?
Unlike the national data, which come from many sources, the state data are compiled from two sources: The Youth Risk Behavior Surveillance Survey (YRBSS) and the CDC Wonder mortality databases (using data collected by National Vital Statistics System). These sources for the state data are the same as those used for the national data (with the exception of marijuana and binge drinking; please see explanation below in “Can national and state data be directly compared?”). As noted under Data Availability, some states choose not to collect data on every objective and some states have their own data sources for some of the 21 Objectives.
How do all the many data sources vary?
These data sources, especially national data sources, vary tremendously in areas including:
1. time frame,
2. age group, and
3. year of baseline and final data collection. (Please view this table for more detail.)
There is less variation among state data sources, where data are drawn only from YRBSS and national vital statistics:
Overall Mortality: 1998 baseline and 2007 final
Motor Vehicle, Homicide and Suicide Mortality: 1999 baseline and 2007 final
Behavioral Objectives (YRBSS): 1999 baseline and 2009 final
Can national and state data be directly compared?
Yes, with the exception of marijuana and binge drinking, the state data are drawn from mortality statistics and YRBSS data and corresponds with the national data. For marijuana and binge drinking, the National Healthy People 2010 data source is the National Survey on Drug Use and Health (NSDUH); the state data source is YRBSS. Because these sources differ, one CANNOT compare numbers (the state numbers will almost always look higher). We can, however, compare changes over time and disparities. For example, it is noteworthy if binge drinking goes up in one state in contrast to a national decline.
Why are safety belt use and condom some shown in the negative, but described in the positive?
In general, YRBSS presents data about the proportion of adolescents engaging in a risky behavior, including rarely or never using a safety belt or not using a condom at last intercourse. By contrast, some of the Healthy People 2010 critical objectives are stated in the positive, such as always, usually or sometimes wearing a seat belt or using a condom. In most cases, the text matches the actual critical objective, even if it is the reverse of how YRBS presents data. The following paragraph describes main exceptions to this rule.
Why are some of the sexual behavior objectives presented in the inverse?
In short, we describe change in safer sexual behavior (Objectives 18a, 18b & 18c) in the manner that makes most intuitive sense. The longer explanation relates to the official Healthy People 2010 measures. When originally developed, these three objectives were combined as one measure: an adolescent was considered safe from sexually transmitted infections and unwanted pregnancy if they abstained from sexual intercourse or if they used a condom. Using the measures available from YRBSS, abstaining is defined as either “never had sex” (18a) or “currently sexually inactive,” (18b) among those who were sexually experienced. Salutary change, therefore, is defined as more teens being sexually inexperienced or more sexually experienced teens being currently sexually inactive. We found it unwieldy and confusing to describe “an increase (or decrease) in the percentage of adolescents who report being sexually inexperienced,” or “an increase (or decrease) in the percentage of sexually experienced adolescents who report being currently sexually inactive.” It was even more confusing to compare subgroups and state and national rates. Therefore, the text presents objective 18.a, sexual inexperience, as being sexually experienced (or having had sexual intercourse); text presents 18.b, currently sexual inactive (among sexually experienced teens), as the percentage who are currently sexually active. Please feel free to contact us, with further questions.
Why are pregnancy data not available for all states and/or not presented in summary text?
The age group for the Healthy People pregnancy objective (09-07) is for ages 15-17. While there are baseline data for this objective (1996), we were unable to locate data for this age grouping for the final year. Pregnancy data for baseline (1996) were collected from the MMWR Report on National and State-Specific Pregnancy Rates Among Adolescents, United States, 1995-1997.