Self-reporting survey methods sometimes lead to serious underestimation of drinking levels in several population groups, according to new findings from a team of Australian researchers. Researchers and public health officials can use a number of survey methods to estimate the level of alcohol consumption in society as a whole and in certain segments of society. Such survey outcomes are vitally important, since accurate tracking of alcohol-related harm requires accurate tracking of alcohol use. In a study published in March 2015 in the Journal of Studies on Alcohol and Drugs, researchers from three Australian institutions compared the accuracy of two self-reporting methods of estimating alcohol intake levels. These researchers concluded that some self-reporting survey methods may lead to a serious underestimation of alcohol consumption, especially in certain portions of the population.
Measuring Alcohol Consumption
Methods used to measure alcohol consumption levels in large or small groups of people go by names that include quantity-frequency measurement, graduated frequency measurement and weekly recall. Quantity-frequency surveys attempt to determine how much alcohol participants drink in a given span of time, as well as how often they drink. Graduated frequency surveys start by ascertaining the maximum amount of alcohol each participant has consumed in a single session, as well as ascertaining how often this maximum level of intake has occurred. Researchers conducting these types of surveys also ask each participant to reveal how often he or she has consumed progressively smaller amounts of alcohol. In a weekly recall survey, researchers ask each participant to estimate how much alcohol he or she has consumed in the very recent past. Not all alcohol consumption surveys gather information in equal detail. While some surveys focus on fairly broad trends in alcohol use, others include data on such things as the specific types of alcohol consumed by each participant and the specific locations in which alcohol use typically occurs. The type of survey used and the level of detail in a survey’s questions can significantly affect the accuracy of the responses given by participants, and all survey methods have their acknowledged limitations.
Alcohol Surveys in the U.S.
Public health officials in the U.S. use several large-scale survey projects to record America’s level of involvement in alcohol use and level of exposure to alcohol-related harm. One prominent federally sponsored project is the National Survey on Drug Use and Health, which uses detailed questioning of nationally representative groups of people 12 and older to track the frequency of alcohol use and level of involvement in dangerous practices called binge drinking and heavy drinking. Another prominent federally sponsored project, called Monitoring the Future, focuses on the use of alcohol (and all other well-known mind-altering substances) among teenagers in the eighth, 10th and 12th grades.
Measuring Accuracy
Researchers who investigate alcohol consumption are well-aware that self-reporting of drinking levels can lead to a substantial underestimation of alcohol use trends in any population. In the study published in the Journal of Studies on Alcohol and Drugs, researchers from Australia’s University of New South Wales, Centre for Alcohol Policy Research and Monash University compared the relative accuracy of two self-reporting methods: graduated frequency surveys and more detailed surveys that ask questions about drinking locations and the types of alcohol consumed during drinking sessions. The graduated frequency survey in question was Australia’s National Drug Strategy Household Survey. The location- and alcohol type-specific survey was conducted by Australian representatives of a project called the International Alcohol Control Study. The researchers concluded that, compared to the results of the more detailed International Alcohol Control Study, the graduated frequency-based National Drug Strategy Household Survey underestimated alcohol consumption levels by fully one-third. Specific groups affected by this underestimation of drinking levels include middle-aged women (49 percent), young men (40 percent) and young women (15 percent). The researchers also concluded that the graduated frequency approach may lead to overestimation of alcohol consumption in older women. Interestingly, the study’s authors concluded that people who frequently participate in the dangerous, drunkenness-inducing practice of binge drinking tend to accurately report their alcohol consumption levels more often than people who rarely binge drink or never binge drink. The authors believe that survey designers must devise methods to offset the underestimation of alcohol consumption that occurs for certain segments of the population when graduated frequency measurements are used.
