Menu
Home
People
Places
Arts
History
Plants & Animals
Science
Life & Culture
Technology
Reference.org
Bias (statistics)
open-in-new
<h2 id="bias-of-an-estimator">Bias of an estimator</h2> <p class="note">Main article: <a href="/facts/Bias_of_an_estimator/oxIvEgmd">Bias of an estimator</a></p> <p>Statistical bias is a feature of a <a href="/facts/Statistics/DNPsKGYU">statistical</a> technique or of its results whereby the <a href="/facts/Expected_value/1XV0JKL8">expected value</a> of the results differs from the true underlying quantitative <a href="/facts/Parameter/zFvVFMv9">parameter</a> being <a href="/facts/Estimation_theory/4nbgFDec">estimated</a>. The bias of an estimator of a parameter should not be confused with its degree of precision, as the degree of precision is a measure of the sampling error. The bias is defined as follows: let T {\displaystyle T} be a statistic used to estimate a parameter θ {\displaystyle \theta } , and let E ( T ) {\displaystyle \operatorname {E} (T)} denote the expected value of T {\displaystyle T} . Then, </p> bias ( T , θ ) = bias ( T ) = E ( T ) − θ {\displaystyle \operatorname {bias} (T,\theta )=\operatorname {bias} (T)=\operatorname {E} (T)-\theta } <p>is called the bias of the statistic T {\displaystyle T} (with respect to θ {\displaystyle \theta } ). If bias ( T , θ ) = 0 {\displaystyle \operatorname {bias} (T,\theta )=0} , then T {\displaystyle T} is said to be an <i>unbiased estimator</i> of θ {\displaystyle \theta } ; otherwise, it is said to be a <i>biased estimator</i> of θ {\displaystyle \theta } . </p><p>The bias of a statistic T {\displaystyle T} is always relative to the parameter θ {\displaystyle \theta } it is used to estimate, but the parameter θ {\displaystyle \theta } is often omitted when it is clear from the context what is being estimated. </p> <h2 id="types">Types</h2> <p>Statistical bias comes from all stages of data analysis. The following sources of bias will be listed in each stage separately. </p> <h3>Data selection</h3> <p><a href="/facts/Selection_bias/XmeLAXBF">Selection bias</a> involves individuals being more likely to be selected for study than others, <a href="/facts/Sampling_bias/nrr4qR2T">biasing the sample</a>. This can also be termed selection effect, <a href="/facts/Sampling_bias/nrr4qR2T">sampling bias</a> and <i><a href="/facts/Berkson%2527s_paradox/X5kM0ESL">Berksonian bias</a></i>.<a class="footnote-ref" id="fnref:3" href="#fn:3"><sup>3</sup></a> </p> <ul><li><a href="/facts/Spectrum_bias/559axvlw">Spectrum bias</a> arises from evaluating diagnostic tests on biased patient samples, leading to an overestimate of the <a href="/facts/Sensitivity_and_specificity/pzc9qfdz">sensitivity and specificity</a> of the test. For example, a high prevalence of disease in a study population increases positive predictive values, which will cause a bias between the prediction values and the real ones.<a class="footnote-ref" id="fnref:4" href="#fn:4"><sup>4</sup></a></li> <li><a href="/facts/Selection_bias/XmeLAXBF">Observer selection bias</a> occurs when the evidence presented has been pre-filtered by observers, which is so-called <a href="/facts/Anthropic_Bias_(book)/XXcYhouX">anthropic principle</a>. The data collected is not only filtered by the design of experiment, but also by the necessary precondition that there must be someone doing a study.<a class="footnote-ref" id="fnref:5" href="#fn:5"><sup>5</sup></a> An example is the impact of the Earth in the past. The impact event may cause the extinction of intelligent animals, or there were no intelligent animals at that time. Therefore, some impact events have not been observed, but they may have occurred in the past.<a class="footnote-ref" id="fnref:6" href="#fn:6"><sup>6</sup></a></li> <li>Volunteer bias occurs when volunteers have intrinsically different characteristics from the target population of the study.<a class="footnote-ref" id="fnref:7" href="#fn:7"><sup>7</sup></a> Research has shown that volunteers tend to come from families with higher socioeconomic status.<a class="footnote-ref" id="fnref:8" href="#fn:8"><sup>8</sup></a> Furthermore, another study shows that women are more probable to volunteer for studies than men.<a class="footnote-ref" id="fnref:9" href="#fn:9"><sup>9</sup></a></li> <li><a href="/facts/Funding_bias/qsRhlxLr">Funding bias</a> may lead to the selection of outcomes, test samples, or test procedures that favor a study's financial sponsor.<a class="footnote-ref" id="fnref:10" href="#fn:10"><sup>10</sup></a></li> <li><a href="/facts/Attrition_bias/XmeLAXBF">Attrition bias</a> arises due to a loss of participants, e.g., loss of follow up during a study.<a class="footnote-ref" id="fnref:11" href="#fn:11"><sup>11</sup></a></li> <li><a href="/facts/Recall_bias/9j6N5DVt">Recall bias</a> arises due to differences in the accuracy or completeness of participant recollections of past events; for example, patients cannot recall how many cigarettes they smoked last week exactly, leading to over-estimation or under-estimation.</li></ul> <h3>Hypothesis testing</h3> <p><a href="/facts/Type_I_and_type_II_errors/aazbF3fq">Type I and type II errors</a> in <a href="/facts/Statistical_hypothesis_testing/yv9RPF6U">statistical hypothesis testing</a> leads to wrong results.<a class="footnote-ref" id="fnref:12" href="#fn:12"><sup>12</sup></a> Type I error happens when the null hypothesis is correct but is rejected. For instance, suppose that the null hypothesis is that if the average driving speed limit ranges from 75 to 85 km/h, it is not considered as speeding. On the other hand, if the average speed is not in that range, it is considered speeding. If someone receives a ticket with an average driving speed of 7 km/h, the decision maker has committed a Type I error. In other words, the average driving speed meets the null hypothesis but is rejected. On the contrary, Type II error happens when the null hypothesis is not correct but is accepted. </p><p>Bias in hypothesis testing occurs when the power (the complement of the type II error rate) at some alternative is lower than the supremum of the Type I error rate (which is usually the significance level, α {\displaystyle \alpha } ). Equivalently, if no rejection rate at any alternative is lower than the rejection rate at any point in the null hypothesis set, the test is said to be unbiased.<a class="footnote-ref" id="fnref:13" href="#fn:13"><sup>13</sup></a> </p> <h3>Estimator selection</h3> <p>The <a href="/facts/Bias_of_an_estimator/oxIvEgmd">bias of an estimator</a> is the difference between an estimator's expected value and the true value of the parameter being estimated. Although an unbiased estimator is theoretically preferable to a biased estimator, in practice, biased estimators with small biases are frequently used. A biased estimator may be more useful for several reasons. First, an unbiased estimator may not exist without further assumptions. Second, sometimes an unbiased estimator is hard to compute. Third, a biased estimator may have a lower value of mean squared error. </p> <ul><li>A biased estimator is better than any unbiased estimator arising from the <a href="/facts/Poisson_distribution/CvCzXkHr">Poisson distribution</a>.<a class="footnote-ref" id="fnref:14" href="#fn:14"><sup>14</sup></a><a class="footnote-ref" id="fnref:15" href="#fn:15"><sup>15</sup></a> The value of a biased estimator is always positive and the mean squared error of it is smaller than the unbiased one, which makes the biased estimator be more accurate.</li></ul> <ul><li><a href="/facts/Omitted-variable_bias/eoDm4AJX">Omitted-variable bias</a> is the bias that appears in estimates of parameters in regression analysis when the assumed specification omits an independent variable that should be in the model.</li></ul> <h3>Analysis methods</h3> <ul><li> Detection bias occurs when a phenomenon is more likely to be observed for a particular set of study subjects. For instance, the <a href="/facts/Syndemic/2eKbeWX6">syndemic</a> involving <a href="/facts/Obesity/RBF9y5Vl">obesity</a> and <a href="/facts/Diabetes_mellitus/2xeHXbsI">diabetes</a> may mean doctors are more likely to look for diabetes in obese patients than in thinner patients, leading to an inflation in diabetes among obese patients because of skewed detection efforts.</li> <li>In <a href="/facts/Educational_measurement/2BGXqEkK">educational measurement</a>, bias is defined as "Systematic errors in test content, test administration, and/or scoring procedures that can cause some test takers to get either lower or higher scores than their true ability would merit."<a class="footnote-ref" id="fnref:16" href="#fn:16"><sup>16</sup></a> The source of the bias is irrelevant to the trait the test is intended to measure.</li> <li><a href="/facts/Observer_bias/BYk0H6p6">Observer bias</a> arises when the researcher subconsciously influences the experiment due to <a href="/facts/Cognitive_bias/49YKVhmN">cognitive bias</a> where judgment may alter how an experiment is carried out / how results are recorded.</li></ul> <h3>Interpretation</h3> <p><a href="/facts/Reporting_bias/nXgoRwTu">Reporting bias</a> involves a skew in the availability of data, such that observations of a certain kind are more likely to be reported. </p> <h2 id="addressing-statistical-bias">Addressing statistical bias</h2> <p>Depending on the type of bias present, researchers and analysts can take different steps to reduce bias on a data set. All types of bias mentioned above have corresponding measures which can be taken to reduce or eliminate their impacts. </p><p>Bias should be accounted for at every step of the data collection process, beginning with clearly defined research parameters and consideration of the team who will be conducting the research.<a class="footnote-ref" id="fnref:17" href="#fn:17"><sup>17</sup></a> <a href="/facts/Observer_bias/BYk0H6p6">Observer bias</a> may be reduced by implementing a <a href="/facts/Blind_study/6ALSxX8M">blind</a> or <a href="/facts/Double-blind/6ALSxX8M">double-blind</a> technique. Avoidance of <a href="/facts/P-Hacking/X11G2jBn">p-hacking</a> is essential to the process of accurate data collection. One way to check for bias in results after is rerunning analyses with different independent variables to observe whether a given phenomenon still occurs in dependent variables.<a class="footnote-ref" id="fnref:18" href="#fn:18"><sup>18</sup></a> Careful use of language in reporting can reduce misleading phrases, such as discussion of a result "approaching" statistical significant as compared to actually achieving it.<a class="footnote-ref" id="fnref:19" href="#fn:19"><sup>19</sup></a> </p> <h2 id="see-also">See also</h2> <ul><li><a href="/facts/Trueness/JQ47xXDq">Trueness</a></li> <li><a href="/facts/Systematic_error/q7pa7bRt">Systematic error</a></li></ul> <h2 id="external-links">External links</h2> <ul><li><a href="https://catalogofbias.org/">The Catalogue of Bias</a> is a project at the <a href="/facts/Centre_for_Evidence-Based_Medicine/GFJ36jvf">Centre for Evidence-Based Medicine</a> cataloguing biases that affect health evidence.</li></ul> <h2 id="references">References</h2> <ol> <li id="fn:1"><p>Cole, Nancy S. (October 1981). "Bias in testing". American Psychologist. 36 (10): 1067–1077. doi:10.1037/0003-066X.36.10.1067. ISSN 1935-990X. <a href="http://doi.apa.org/getdoi.cfm?doi=10.1037/0003-066X.36.10.1067" target="_blank">http://doi.apa.org/getdoi.cfm?doi=10.1037/0003-066X.36.10.1067</a> <a href="#fnref:1" class="footnote-back-ref">↩</a></p></li> <li id="fn:2"><p>Popovic, Aleksandar; Huecker, Martin R. (June 23, 2023). "Study Bias". Stat Pearls. PMID 34662027. <a href="https://www.ncbi.nlm.nih.gov/books/NBK574513/" target="_blank">https://www.ncbi.nlm.nih.gov/books/NBK574513/</a> <a href="#fnref:2" class="footnote-back-ref">↩</a></p></li> <li id="fn:3"><p>Rothman, Kenneth J.; Greenland, Sander; Lash, Timothy L. (2008). Modern Epidemiology. Lippincott Williams & Wilkins. pp. 134–137. <a href="/wiki/Kenneth_Rothman_(epidemiologist)" target="_blank">/wiki/Kenneth_Rothman_(epidemiologist)</a> <a href="#fnref:3" class="footnote-back-ref">↩</a></p></li> <li id="fn:4"><p>Mulherin, Stephanie A.; Miller, William C. (2002-10-01). "Spectrum bias or spectrum effect? Subgroup variation in diagnostic test evaluation". Annals of Internal Medicine. 137 (7): 598–602. doi:10.7326/0003-4819-137-7-200210010-00011. ISSN 1539-3704. PMID 12353947. S2CID 35752032. <a href="https://pubmed.ncbi.nlm.nih.gov/12353947/" target="_blank">https://pubmed.ncbi.nlm.nih.gov/12353947/</a> <a href="#fnref:4" class="footnote-back-ref">↩</a></p></li> <li id="fn:5"><p>Bostrom, Nick (2013-05-31). Anthropic Bias: Observation Selection Effects in Science and Philosophy. New York: Routledge. doi:10.4324/9780203953464. ISBN 978-0-203-95346-4. <a href="978-0-203-95346-4" target="_blank">978-0-203-95346-4</a> <a href="#fnref:5" class="footnote-back-ref">↩</a></p></li> <li id="fn:6"><p>Ćirković, Milan M.; Sandberg, Anders; Bostrom, Nick (2010). "Anthropic Shadow: Observation Selection Effects and Human Extinction Risks". Risk Analysis. 30 (10): 1495–1506. Bibcode:2010RiskA..30.1495C. doi:10.1111/j.1539-6924.2010.01460.x. ISSN 1539-6924. PMID 20626690. S2CID 6485564. <a href="https://onlinelibrary.wiley.com/doi/abs/10.1111/j.1539-6924.2010.01460.x" target="_blank">https://onlinelibrary.wiley.com/doi/abs/10.1111/j.1539-6924.2010.01460.x</a> <a href="#fnref:6" class="footnote-back-ref">↩</a></p></li> <li id="fn:7"><p>Tripepi, Giovanni; Jager, Kitty J.; Dekker, Friedo W.; Zoccali, Carmine (2010). "Selection Bias and Information Bias in Clinical Research". Nephron Clinical Practice. 115 (2): c94 – c99. doi:10.1159/000312871. ISSN 1660-2110. PMID 20407272. S2CID 18856450. <a href="https://www.karger.com/Article/FullText/312871" target="_blank">https://www.karger.com/Article/FullText/312871</a> <a href="#fnref:7" class="footnote-back-ref">↩</a></p></li> <li id="fn:8"><p>"Volunteer bias". Catalog of Bias. 2017-11-17. Retrieved 2021-12-18. <a href="https://catalogofbias.org/biases/volunteer-bias/" target="_blank">https://catalogofbias.org/biases/volunteer-bias/</a> <a href="#fnref:8" class="footnote-back-ref">↩</a></p></li> <li id="fn:9"><p>Alex, Evans (2020). "Why Do Women Volunteer More Than Men?". Retrieved 2021-12-22. <a href="https://teamkinetic.co.uk/blog/2019/07/10/women-volunteer-more-than-men/" target="_blank">https://teamkinetic.co.uk/blog/2019/07/10/women-volunteer-more-than-men/</a> <a href="#fnref:9" class="footnote-back-ref">↩</a></p></li> <li id="fn:10"><p>Krimsky, Sheldon (2013-07-01). "Do Financial Conflicts of Interest Bias Research?: An Inquiry into the "Funding Effect" Hypothesis". Science, Technology, & Human Values. 38 (4): 566–587. doi:10.1177/0162243912456271. ISSN 0162-2439. S2CID 42598982. <a href="https://doi.org/10.1177/0162243912456271" target="_blank">https://doi.org/10.1177/0162243912456271</a> <a href="#fnref:10" class="footnote-back-ref">↩</a></p></li> <li id="fn:11"><p>Higgins, Julian P. T.; Green, Sally (March 2011). "8. Introduction to sources of bias in clinical trials". In Higgins, Julian P. T.; et al. (eds.). Cochrane Handbook for Systematic Reviews of Interventions (version 5.1). The Cochrane Collaboration. <a href="/wiki/Julian_Higgins" target="_blank">/wiki/Julian_Higgins</a> <a href="#fnref:11" class="footnote-back-ref">↩</a></p></li> <li id="fn:12"><p>Neyman, Jerzy; Pearson, Egon S. (1936). "Contributions to the theory of testing statistical hypotheses". Statistical Research Memoirs. 1: 1–37. <a href="/wiki/Jerzy_Neyman" target="_blank">/wiki/Jerzy_Neyman</a> <a href="#fnref:12" class="footnote-back-ref">↩</a></p></li> <li id="fn:13"><p>Casella, George; Berger, Roger L. (2002), Statistical Inference, 2nd Ed., p387 <a href="#fnref:13" class="footnote-back-ref">↩</a></p></li> <li id="fn:14"><p>Romano, Joseph P.; Siegel, A. F. (1986-06-01). Counterexamples in Probability And Statistics. CRC Press. pp. 194–196. ISBN 978-0-412-98901-8. <a href="978-0-412-98901-8" target="_blank">978-0-412-98901-8</a> <a href="#fnref:14" class="footnote-back-ref">↩</a></p></li> <li id="fn:15"><p>Hardy, Michael (2003). "An Illuminating Counterexample". The American Mathematical Monthly. 110 (3): 234–238. doi:10.2307/3647938. ISSN 0002-9890. JSTOR 3647938. <a href="https://www.jstor.org/stable/3647938" target="_blank">https://www.jstor.org/stable/3647938</a> <a href="#fnref:15" class="footnote-back-ref">↩</a></p></li> <li id="fn:16"><p>National Council on Measurement in Education (NCME). "NCME Assessment Glossary". Archived from the original on 2017-07-22. <a href="/wiki/National_Council_on_Measurement_in_Education" target="_blank">/wiki/National_Council_on_Measurement_in_Education</a> <a href="#fnref:16" class="footnote-back-ref">↩</a></p></li> <li id="fn:17"><p>Popovic, Aleksandar; Huecker, Martin R. (June 23, 2023). "Study Bias". Stat Pearls. PMID 34662027. <a href="https://www.ncbi.nlm.nih.gov/books/NBK574513/" target="_blank">https://www.ncbi.nlm.nih.gov/books/NBK574513/</a> <a href="#fnref:17" class="footnote-back-ref">↩</a></p></li> <li id="fn:18"><p>"5 Types of Statistical Biases to Avoid in Your Analyses". Business Insights Blog. 2017-06-13. Retrieved 2023-08-16. <a href="https://online.hbs.edu/blog/post/types-of-statistical-bias" target="_blank">https://online.hbs.edu/blog/post/types-of-statistical-bias</a> <a href="#fnref:18" class="footnote-back-ref">↩</a></p></li> <li id="fn:19"><p>Popovic, Aleksandar; Huecker, Martin R. (June 23, 2023). "Study Bias". Stat Pearls. PMID 34662027. <a href="https://www.ncbi.nlm.nih.gov/books/NBK574513/" target="_blank">https://www.ncbi.nlm.nih.gov/books/NBK574513/</a> <a href="#fnref:19" class="footnote-back-ref">↩</a></p></li> </ol>