Bias is technically the deviation of the probable value of a chance variable from the resultant correct or consigned value (US NRC, 2007). It is the differentiation involving the experimental average of measurements held at repeated case and a reference value, or referred to as accuracy. Bias is calculated and articulated at a solitary position within the working array of the measurement system (Measurement Media, 2008) and is evident in the gathering of Retrospective accounts.
Retrospective data are acquired through interviews and questionnaires. Conversely, prospective data are attained through the use of existing records taken from previous studies (ABC, 2008). There are three main divisions of bias. The first division is Selection Bias. Basically, selection bias takes place when the topics studied do not give proper body or representation of the target population about which end results are to be taken from (Coggon, Rose, & Barker, 1997).
In selection bias, when the involvement of exposure and alcoholism is dissimilar for those who finish a study evaluated with those who match the characteristics of the target population, the general population is selected; they are for which the measure of effect is being considered (Ibrahim, Alexander, Shy & Farr, 1999). In a case study involving alcoholics, selection bias is characterized where those who volunteer to answer questionnaires may possess unlike character than the proposed individual in the target population. In the main, individuals who do not react to requests to be evaluated have different characteristics than responders.
Bias will be established if the association between exposure and alcoholism differs between the results for the study volunteers and non-responders. The second division of bias is the Information Bias. This major type of bias comes to pass from errors in measuring exposure or alcoholism. In a study to calculate approximately the relative risk of alcohol intake and road accidents, associated with exposure to wines, beers and spirits, alcoholics were solicited for information about their contact and exposure with such substances before driving, and their responses were compared with those from control alcoholics.
With this devise, there is a hazard that “case,” or variable, mothers, who are extremely goaded to find out what they drank the most in the expanse of the drinking session, might recollect memories of past contact more completely than the alcoholics from the control group. If that would be the case, a bias would product with a propensity to overstate risk estimates (Coggon, Rose, & Barker, 1997). Recall Bias is included in this type of bias. Recall bias happens when a respondent is asked to relate to a particular topic, and they either exaggerate what information or rule out information they think isn’t appropriate for inclusion.
Data could be inconsistent or flawed when epidemiological study results are deduced via retrospective data gathering (ABC, 2008). In the case of the alcoholics, recall bias might prove to be a threat. If an accident happens, and excessive alcohol consumption is taken as the culprit, the respondent might give out information on his account rendering a holier than thou rendition of what really happened before. Some might not say that they had been drinking before driving, while others might say that they had alcohol intake more than what was required.
What I see in this is that these people are trying to protect their dignity as a person, and their revelations might be put up against their wills. Seeing as this is a threat, there might be a risk of imbalanced information and results gathered for this particular epidemiological study. In the planning of case studies, the evaluators must see to it that they include the recall bias of their respondents toward a particular topic, especially if they are employing a primary type of data gathering; using questionnaires, surveys and interviews.
References 74. 4 Definitions. (2007). Online, United States Nuclear Regulatory Commission. Retrieved on July 10, 2008 from http://www. nrc. gov/reading-rm/doc-collections/cfr/part074/part074-0004. html Bias. (2008). Online, Measurement Databases for Industry and Science. Retrieved on July 10, 2008 from http://measurementdb. com/index. php? name=Sections&req=viewarticle&artid=17&page=1 Coggon, D. , Rose, G. & Barker, DJP (1997). Epidemiology for the Uninitiated. Online, BMJ Publishing. Retrieved on July 10, 2008 from http://www.
bmj. com/epidem/epid. 4. html Hassan, E. (2006). Recall bias can be a threat to retrospective and prospective research designs. Internet Journal of Epidemiology, 2(3), 4-4. Ibrahim, M. Alexander, L. Shy, C. & Farr, S. (1999). Selection Bias. PDF File, ERIC Notebook. Retrieved on July 10, 2008 from http://www. durham. hsrd. research. va. gov/eric/notebook/ERICIssue08. pdf What is Recall Bias? (2008). Online, Abortion Breast Cancer (ABC). Retrieved on July 10, 2008 from http://www. abortionbreastcancer. com/bias/index. htm