On the surface, age seems easy enough to measure, and it is for many research projects. However, the concept of age can vary depending on a variety of factors, including the purpose of the research, the target respondents, policy implications, and such. For example, a research project that measures demographics of a broad population may measure age in one-year increments. Another project may group age based on some predefined ordinal scale. Yet another project may be interested in a particular age group that may be tied to other foci of the research, such as at risk populations, adolescents, elderly, developmental and lifestyle issues, etc. (Chaffee, 1991).
Typically, age is conceptualized as the length of time, most often in completed years, that a given person has been alive, measured at the beginning of birth. It is very often measured as a ratio variable, which has the important statistical advantage over other levels of measurement in that it captures variance others may miss (e.g., measuring age as an ordinal variable does not distinguish among ages within a measured range). Thus, asking someone how old they are will likely obtain responses in years, expressed as positive integers. When measured as a continuous variable, the distribution of age in a sample of the general population will certainly be skewed right since there are fewer yet extreme older people than younger (you can't be younger than 0). A common ordinal measure of age might ask respondent to "stop" once they hear the interviewer reach their age range, such as 18-24, 25-34, 35-44, 45-54, 55-64, older than 64. Notice that this scale treats age as an ordinal variable since all categories do not contain equal age ranges. Technically, parametric statistics should not be performed on such a scale, although it does occur.
Some research that focuses on children and adolescents merely obtain school grade levels, which can serve as a surrogate for age, since the ages of most students in the same grade are within a year or two of each other. However, the definition of "adolescence" varies across research project and across health agencies and foundations. For example, the Centers for Disease Control (CDC) define adolescence and young adult from age 10 to 24, with the complete age range defined as "youth" (CDC, 2010). Defined age categories of children, youth, and teens can fall within that age range. The Maternal Child Health Bureau (MCHB, DHHS) defines adolescence from age 11 to age 21. The definition used by Bright Futures Guidelines for Health Supervision also includes children from age 11 through age 21. The 2007 National Survey on Drug Use and Health (NSDUH) defined "youth" as ages 12-17, young adults as ages 18-25, and adults 26+.
Significant research is also conducted on older adults as well. In 2008 and 2009, CDC, AARP, and the American Medical Association collaborated to develop a report highlighting opportunities to broaden the use of potentially lifesaving preventive services, specifically for adults aged 50-64. The National Institute for Aging is generally interested in adults 65+, although aging issues may be relevant to younger groups.
Regardless of how age is defined and measured, many agencies report age in ordinal age ranges rather than as means with standard deviations. For example, data summarized in the National Health Interview Survey (NHIS) No. 249 breaks age into the following categories: 18—44, 45—64, 65—74 and 75 years and over. This breakout may be marginally useful for readers interested in particular age groups.
The U.S. Census Bureau, which has measured the age of each respondent since 1800, asks people to report their age on the paper form with the item:
What is person 1's age and what is person 1's date of birth? (Please report babies as age 0 when the child is less than 1 year old).
Age on April 1, 2010 __________ Print numbers in boxes. Month:__________ Day: __________ Year:__________
Asking for birth month, day, and year can serve as validation for the age reported on April, 2010.
The Health Information National Trends Surveys (HINTS) of 2003, 2005, and 2007 ask respondents for their age, with small variations across waves.
What is your age? (2003)
May I please have your age? (2005)
Because these questions are very important, I need to ask [again] what is your age, please? (2007)
In each case, interviewers may probe non-responses by asking for these age categories:
less than 18 years old, between 18 and 34; 35 to 39, 40 to 44, 45 or older
In their survey of Caucasian and African American women, Leshner, Cheng, Song, Choi, & Frisby (2007) measured age with the single question:
How old were you on your last birthday?
For many research purposes, the best strategy to measure age is to obtain age as the ratio variable completed years since birth. Data for age can later be collapsed into age categories that are suitable to particular analyses and reporting. However, measuring age as an ordinal variable cannot be later converted to a ratio measure. In most research circumstances, it is better to use the most precise measure as long as it does not pose an unnecessary hardship on the respondent.
Centers for Disease Control and Prevention, National Center for Injury Prevention and Control. Web-based Injury Statistics Query and Reporting System (WISQARS) [online]. (2010) [cited 2010 June 14]. Available from: URL: www.cdc.gov/injury/wisqars/index.html youth=10-24
Chaffee, S.H. (1991).Communication Concepts
1: Explication. Newbury Park, CA: Sage.
Leshner, G., Cheng, I., Song, H., Choi, Y., & Frisby, C. (2006).The role of spiritual health locus of control in breast cancer information processing between African American and Caucasian women.
Integrative Medicine Insights, 2, 35-44.
Substance Abuse and Mental Health Services Administration. (2008).Results from the 2007 National Survey on Drug Use and Health: National Findings
(Office of Applied Studies, NSDUH Series H-34, DHHS Publication No. SMA 08-4343). Rockville, MD.
U.S. Census Bureau. (2010). The Questions on the Form. Retrieved May 20, from http://2010.census.gov/2010census/how/interactive-form.php