Recall bias is a systematic deviation from the truth, which occurs when the respondent’s answers are affected by his or her memory (Hassan, 2006). Recall bias is one of the factors that affect the credibility or validity of the research. In epidemiological studies, the concept of recall bias is equally important. This is because the correctness of the respondent’s answer will determine the level of dependence one can place in such research. The correctness of the answer is usually dependent on the memory of the respondent. Therefore, this article analyses the concept of recall bias in epidemiological studies involving alcohol consumption.
Various factors can contribute to recall bias in research. Anders et al in their paper ‘A Randomized Study of Secondary Prevention of Early Stage Problem Drinkers in Primary Health Care’ introduced a few factors in the methods that could lead to recall bias (1989). The element of time used in this research could introduce recall bias in the study. The questionnaires were delivered using mail and therefore, it took some time before they reached the respondents.
To add to this, the questionnaire focused on the previous week’s intake, during the weekdays and the weekend (Anders 1989). The time between the consumption of alcohol and answering the questionnaire is considerably long. During this time, the memory of the respondents may be distorted or may change thus introducing recall bias in the methods. Moreover, the researchers measured the alcohol consumption, and habits on basis of the previous week’s intake.
Time factor tends to affect the memory of the respondents. Each of the respondents will give an account of his or her alcohol consumption habits depending on what he or she recalls. This may reduce the level of accuracy in measurements because the respondent’s memory may not be very sharp after some time has passed. Therefore, to enhance this research, immediate analysis is required to ensure that the concept of recall bias is minimized.
Miles & Strang introduce an element of recall bias in the methods they used to analyze alcohol consumption and addiction (2001). The authors provided questionnaires to young people who had to report on their consumption habits. The subjects of the study had to report on the number of units consumed in the previous month and the number of units they consume on a typical day (Miles & Strang, 2001). Moreover, the questionnaire seeks to record data on the frequency of alcohol consumption. For example, data on weekly consumption, daily consumption, and monthly consumption. To add to this the researchers aim to find out the date that the subjects began drinking.
All these elements may contribute to recall bias. This is because when the subject is answering questions on daily alcohol consumption the information will be very accurate since his or her memory is not distorted by the time factor. However, when the subjects answer questions about monthly and weekly consumption, there may be deviations in the data given. This is because the period between consumption and the time of actually answering the questionnaire is long.
During this time, the subject may confuse or lose some memory about his or her consumption habits thus distorting information. The element of recall bias in this study may interfere with the results obtained by the researchers. Therefore, researchers have to introduce some form of control to ensure that the bias introduced in the study is eliminated. In the case of the time factor, researchers should perform their questioning immediately after consumption.
References
Anders, R., Lena, A., Hakan, B., Stefan B., Claes G., Olle H., Annika Hassler, Anna K., Peter M., Roland M., Kjell Nyman, Anders O., Eva O., Anders R., & Olle W. A. (1989). Randomized Study of Secondary Prevention of Early Stage Problem Drinkers in Primary Health Care. British Journal of Addiction, 84(1), 1319-1327.
Hassan, E. (2006). Recall Bias can be a Threat to Retrospective and Prospective Research Designs. Internet Journal of Epidemiology, 3(2), 1-10.
Miles, H. Winstock, A. & Strang, J. (2001). Identifying young people who drink too much: the clinical utility of the five-item Alcohol Use Disorders Identification Test (AUDIT). Drug and Alcohol Review, 20 (1), 9-18.