Data analysis is vital for most if not all business decisions. Businesses are increasingly using data in everyday predictions and strategic decision making. Even with the most sophisticated procedures put in place to ensure credible analysis, most companies are liable to favoritism and unfairness. Being bias occurs naturally and we all exhibit it one way or another. What can be done to minimize bias? I will be exploring the different types of bias and how to minimize and possibly overcome their effect on the outcome.
Confirmation bias arises when the researcher chooses accept feedbacks that support the researcher’s view of the cited situation. The researcher is not open to possibilities that oppose his/her description. These biased search processes lead to the maintenance of the information seeker’s position, even if the position is not justified on the basis of all available information .According to dissonance theory (Festinger, 1957) people prefer information that support their argument than those that are opposed to them. This type of thinking makes the company conform to a fixed way of thinking. This is dangerous particularly to a company that is dependent on varied feedbacks for example a hotel.
Delegate your assignment to our experts and they will do the rest.
Confirmation bias often occurs when the researcher is not willing to consider opposing facts. As such, one solution to this is to require the researcher to be open to opposing positions and look at them critically and not being blinded to his/her preconceived position and accept in cases where the opposing argument is justified. Triangulation can be used to ensure objectivity by using two or more methods or sources before coming to a conclusion on a research.
In researches where an individual selects the subjects to be examined instead of choosing at random, there is likely to be selection bias. The researcher may pick subjects using two or more opposing categories (for example male or female) but the individuals might be exposed to similar circumstances making the outcome biased. If potential observation from a population of interest are excluded from a sample on a non-random basis, one risks sample selection bias. ( Berk, 1983 ). Selection bias can also occur when the individuals to be researched select themselves. For example if a researcher is interested in successful entrepreneurs, less successful entrepreneurs might offer themselves to be studied when more successful entrepreneurs may not probably because they are too busy or too hard to reach.
Survivorship bias an example of selection bias arises when the researcher focuses on the subjects of the research that passed a process and totally ignoring those who did not because they may not be there. This causes a wrong conclusion by using only the account of those that survived. For example if five friends from the same high school are the top students in college, it is possible to conclude that the high school might have overall excellent academic performance. This, however, is inconclusive, unless the researcher looks into the performance of other students from the same high school. A sure way of curbing selection bias is by examining and including all available elements or groups for research not just the ones the researcher feels meet a criteria. Excluding key subjects may lead to wrong conclusions of the study.
Funding bias is the tendency of research outcomes to favor the views of the sponsors or the company. This undermines its legitimacy. The researcher might be compelled to or have no other choice but to give a positive review of a research as the truth could affect the funding and rating of the company. Industry also provides individual physicians or entire academic departments to with unrestricted funds to that can be applied towards personal or institutional research initiatives. These forms of compensation may undermine investigator’s objectivity by rewarding those who produce results most favorable to the sponsor’s interest.
A solution to funding bias would be to ask the respective companies and institutions to take full obligation for the outcome of the researches they carry out. To ensure objectivity in clinical research, some investigators have suggested that the industry-academia collaborations continue only if academic medical centers assume sole responsibility for the design, conduct, analysis and reporting of clinical trials. Others have supported the creation of conflict-of-interest committees at academic institutions to monitor the financial interests of both clinician-investigators and institutional decision makers ( Johns, Barnes & Florencio, 2003).
References
Berk, R. A. (1983). An introduction to sample selection bias in sociological data. American Sociological Review , 386-398.
Chopra, S. S. (2003). Industry funding of clinical trials: benefit or bias?. Jama , 290 (1), 113-114.
Jonas, E., Schulz-Hardt, S., Frey, D., & Thelen, N. (2001). Confirmation bias in sequential information search after preliminary decisions: an expansion of dissonance theoretical research on selective exposure to information. Journal of personality and social psychology , 80 (4), 557.
Johns, M. M., Barnes, M., & Florencio, P. S. (2003). Restoring balance to industry-academia relationships in an era of institutional financial conflicts of interest: promoting research while maintaining trust. Jama , 289 (6), 741-746.
