Three biases that can influence the outcome of Data Analysis
Verified
Added on  2023/04/24
|7
|1529
|134
AI Summary
This report will provide an insight upon three biases namely cognitive, inter-group and conformation biases that influences outcomes of data analysis, what are they and how such biases are caused. The report concludes after making suggestions on how these biases can be minimised to avoid errors in data analysis.
Contribute Materials
Your contribution can guide someone’s learning journey. Share your
documents today.
1 Title of Paper Student Name Course/Number Due Date Faculty Name
Secure Best Marks with AI Grader
Need help grading? Try our AI Grader for instant feedback on your assignments.
2 Three biases that can influence the outcome of Data Analysis Every research paper consists of significant data analysis that are designed, steered and presented in transparent and understandable manner, without getting eccentric from the truth. Data analysis that do not complaint with primary objectives is considered to be misleading. Nevertheless, some analysis creates false conclusions or distorted results which gives wrong outcomes, thereby resulting in substantial loses in businesses or other decision- making outcomes(Friedman, 2017). This report will provide an insight upon three biases namely cognitive, inter-group and conformation biases that influences outcomes of data analysis, what are they and how such biases are caused. This report will conclude after making suggestions on how these biases can be minimised to avoid errors in data analysis. Definition of bias in data analysis Bias is deviation from truth in data analysis, data collection and data interpretation that creates false conclusions. Biases can take place either unintentionally or with an intention to introduce bias in someone’s data analysis. Information providers introduce biasness in data analysis for making their analysed data look more preferable and in favour of research hypothesis. There are many ways through which bias decision can be implemented during data analysis including fabricating, manipulating or abusing which results in reporting non- existing data and inappropriate statistics in front of its readers (Montibeller & Winterfeldt, 2015, p. 1231). Three types of biases influencing data analysis outcomes Cognitive bias is one of those biases that follows a systematic pattern deviating from normal and rational judgement, thereby way of thinking often resulting in illogical and holistic influence. This mental bias can cause an individual create his/her own kind of social
3 or work-related authenticity, an objective reality created from individual perception. Moreover, the construction of reality world is never objective that results in misleading individuals towards irrational behaviour(Montibeller & Winterfeldt, 2015). Cognitive bias often leads to perceptual distortion, illogical interpretation and inaccurateoutcomes of data analysed. Cognitive bias can arise due to several reasons during data analysis procedure including mental noise, information processing shortcuts, emotional motivations and limited information gathering. In literature, it is apparent that people do not behave rationally every time. There are times when they are presented with more information than required thereby trying to make quick decisions. This leads them rely on cognitive shortcuts, also known as heuristics (Friedman, 2017). To minimise or avoid such biases, a detailed study of data analysis can be made especially in the areas where changes have been made from the start.Change affecting the entire decision-making process can be noted along with making out which metrics have been changed purposely. This method can also be called as relational impact under which reasons behind impacts over decisions can be found out to clarify data interpretation. Rational analysis method can also be applied to make decisions after deficiency in decisions is observed(Ehrlinger , Readinger, & Kim, 2016). Inter-group bias is one of the biases that can take in the form of ingroup liking like distribution of rewards among group members or protecting ingroup from negative results. Although group standards can be evaluated on the basis of individual standards and goodness, regardless of outgroups, most of them remains unclear thereby resulting in arousing social comparison and misinterpretation of information’s and data. Apparently, comparison always brings unhealthy work environment but the real issue arises when comparisons are made unethically. Realistic conflicts theory suggests that inter-group biases arise from competitions for resources among group members. Since each individual compete for same resource, they
4 serve best interests to favour group members rather than spurning outsiders. Such kind of behaviour may sometimes result in more hostile situations along with distorting available data due to intense rivalry(Steffens , Reese, Ehrke, & Jonas, 2017). Social categorization can contribute in minimising inter-group biases effectively. It will not only motivate elimination of ethnic violence, but also make complex data simpler through joint efforts of unbiased group activities. Secondly, the paradigmatic evidences of cultural and ethnic violence can be equated with its outcomes and further analysed from disciplinary perspective. Inter-group bias is not arbitrary due to which it can be controlled easily. To avoid such biases, it is best to remain informed and aware from the beginning of data analysis process so that thinking objectively becomes possible even though bias threatens rationality during gathering of information’s(Steffens , Reese, Ehrke, & Jonas, 2017). Confirmation bias takes place when individuals makes decisions according to their own beliefs and desires. When people start assuming the idea or any data to be true, they end up believing it and gets motivated by wishful assumptions. This leads individual stop gathering enough information’s and knowledge and once the gathered evidences confirm so far true, they believe it to be true. Once the data is embraced while ignoring or rejecting other significant information, data analysis process casts several doubts in it (Calikli, Arslan, & Berner, 2010, para. 1). The reasons behind arise of confirmation bias can be due to over anxious individuals who interprets information to be dangerous. For instance, people with low self-esteem can prove highly sensitive due to being ignored by other and thus constantly shows signs for which people stop liking them. Confirmation bias may not always arise due to data fluctuation, it may be persistent due to another particular hypothesis. Nevertheless, potential explanation can be identified to seek evidence which supports the explanation behind formulating such decisions(Gatlin, Hallock, & Cooley, 2017).
Paraphrase This Document
Need a fresh take? Get an instant paraphrase of this document with our AI Paraphraser
5 To minimise conformation bias, treatment done at initial stages of information gathering can help in making data analysis without any fluctuations and complete data set after being reviewed for considering reasons behind data validity. Brainstorming can be done, if possible, to identify potential causes behind unexpected fluctuations or distorting of information. However, developing many initial hypotheses may cause development of constraints during decision making process and therefore, using analytical procedures while developing data analysis methods can be recommended(Luippold, Perreault, & Wainberg, 2015). Conclusion Complex data analysis consumes huge portion of information thereby increases problems in effective outcomes of analysis. Such problems can not be solved easily or by simple procedures. To remain effective, decision making must take into account every characteristic involved in data analysis procedure along with interactions conducted to define data analysis issues. The above report states that no matter how severe the process is, the definition of every issue can be represented by the nature of issue. The principles described above regarding biases influencing outcomes of analysis suggests that every data analysis and decision-making process must contain explicit examination including challenges behind assumptions and biases, underpinning overall data analysis process to mitigate the negative effects of biased orientations.
6 References Calikli, G., Arslan, B., & Bener, A. (2010).PPIG, Universidad Carlos III de Madrid, 2010 www.ppig.org Confirmation Bias in Software Development and Testing: An Analysis of the Effects of Company Size, Experience and Reasoning Skills. Retrieved from https://www.researchgate.net/publication/235430372_Confirmation_Bias_in_Softwar e_Development_and_Testing_An_Analysis_of_the_Effects_of_Company_Size_Expe rience_and_Reasoning_Skills Ehrlinger , J., Readinger, W. O., & Kim, B. (2016).Decision-Making and Cognitive Biases. Retrieved from https://www.researchgate.net/publication/301662722_Decision- Making_and_Cognitive_Biases Friedman, H. H. (2017). Cognitive Biases that Interfere with Critical Thinking and Scientific Reasoning: A Course Module .SSRN Electronic Journal, 1-60. Gatlin, K. P., Hallock, D., & Cooley, L. G. (2017). Confirmation Bias among Business Students: the Impact on Decision-Making.Review of Contemporary Business Research, 06(2), 10-15. Luippold, B. L., Perreault, S., & Wainberg, J. (2015).5 ways to overcome confirmation bias. Retrieved from https://www.journalofaccountancy.com/issues/2015/feb/how-to- overcome-confirmation-bias.html Montibeller, G., & Winterfeldt, D. V. (2015). Cognitive and Motivational Biases in Decision and Risk Analysis.Risk Analysis, 35(07), 1230-1251. Steffens , M. C., Reese, G., Ehrke, F., & Jonas, K. J. (2017).When does activating diversity alleviate, when does it increase intergroup bias? An ingroup projection perspective.
7 Retrieved from https://journals.plos.org/plosone/article?id=10.1371/journal.pone.0178738