Confronting Pattern Recognition and Embracing Diversity
In business and many industries, we create patterns to reduce errors and standardize production. However this same behavior, when taken too far can create unintended negative consequences such as bias. As we hire new employees, invest in new businesses, and offer promotions, how do we ensure that we are making impactful decisions that impact profit, innovation, efficiency, diversity and inclusion? Do we have tendencies, stereotypes, and beliefs that reinforce non-inclusive workplaces? Research indicates that everyone has their own set of unspoken beliefs called the “unconscious bias” (Hewett, 2019). These biases support the use of what Denise Hewett calls “pattern matching” (Hewett, 2019) which reinforces race, gender, educational, and economic disparities in the workplace. In this Insight article, we seek to understand pattern matching, what causes it, how is it used, and how can we prevent erroneous pattern matching to create a more inclusive and profitable workplace.
“Innovation does not fit a pattern” - Denise Hewett, CEO of Scriptd
What is Pattern Matching?
Pattern Matching is used in both algorithms and human nature. In coding and algorithm creation pattern matching is defined as, “the process of checking whether a specific sequence of characters/tokens/data exists among the given data” (educative, n.d). Humans use pattern matching in decision-making processes in much of the same way. John Wiese states that in human nature “pattern matching is the cognitive process by which your brain connects current sensory stimulation with past experience” (Wiese, 2020). In both forms of pattern matching, already learned information is applied to present situations or data. While this concept seems logically sound, it has its own pitfalls and drawbacks. For example, Stereotypes are an ever-present shortcoming in society. They marginalize individuals to the specific characteristics that are often associated with their race, gender, or other social groups. Oftentimes these stereotypes do not represent the individual but influence first impressions and opinions of the individual. When these stereotypes infiltrate the thought process of human beings and qualifications within algorithms, discriminatory pattern matching sequences can unintentionally take place. Because of this, pattern matching has morphed into the unintentional application of learned stereotypes and biases on employees, job applicants, and business practices in general.
What causes errors in Pattern Matching?
There are two essential components of pattern matching errors: stereotypes and unconscious bias. The combination of these two factors creates the foundation for discriminatory pattern matching.
“Stereotypes lose their power when the world is found to be more complex than the stereotype would suggest…” - Ed Koch, Former United States Representative
Stereotypes, according to Dr. Saul Mcleod, are the “over generalized belief about a particular group or class of people.” (2017). Dr. Mcleod states that stereotyping causes us to believe that an individual “has a whole range of characteristics… that we assume all members of that group have” (2017). An example of a stereotype would be the belief that women perform better in businesses focused on service and health (Hewett, 2019). This stereotype cannot accurately be applied to all women, however, it often is. In her speech, Denise Hewett states that investors are more likely to invest in women-owned businesses that skew towards stereotypically female fields such as health and wellness (Hewett, 2019). However, the existence of a stereotype in a society does not always indicate the use of discriminatory policies for each organization. Stereotypes must be left unconfronted by the individuals making decisions for an organization in order to increase discrimination. These unconfronted stereotypes embed themselves in the unconscious bias of business leaders and investors, further supporting discriminatory pattern matching practices.
Unconscious biases are beliefs subconsciously held by individuals. They are not acknowledged and are learned over time through a “combination of data, knowledge, and experience[s]” (Bryant, 2019). This “learned data” can become convoluted with social stereotypes and personal assumptions about social groups (Bryant, 2019). Katherine Bryant of The Progress Partnership states that the “unconscious bias distorts our perception of risk and causes us to be over-reliant on initial data, to disregard information that contradicts our existing beliefs, and engage in potentially damaging group think” (2019). With over 250 recorded cognitive biases, it is essential that companies create safeguards against the biases (Bryant, 2019). An example of one of these biases is the salary history bias. The Center for American Progress states that “employers’ reliance on salary history in hiring and compensation decisions is a textbook example of structural bias” (Bleiweis, 2021). This bias thrives off of the stereotype that job applicants who were paid more in previous positions are better employees. This view is inaccurate as salaries are often dependent on local cost-of-living standards, and has little relevance to applicant job performance when compared to promotions and supervisor reports. This bias specifically reinforces wage gaps between majority and minority groups, and in many states acting on this bias and asking job applicants, their salary history is now considered an illegal business practice (Bleiweis, 2021). Unconscious biases, like the salary history bias, are fueled by stereotypes and when left unacknowledged further promote discriminatory pattern matching.
How is Pattern Matching Used?
Pattern matching has the potential to influence almost any area of business in which decisions are made. However, there are two specific areas in which discriminatory practices can be inadvertently reinforced with pattern matching: Hiring and Promotion Standards as well as Investment Standards.
When seeking out new employees, reading resumes, and reviewing cover letters it is important to confront the unconscious biases that can influence your judgment. In an attempt to further understand bias in the hiring process, Yale University researchers looked for and analyzed the hiring biases of professionals from “research-intensive” universities (Moss-Racusin et al., 2012). Faux laboratory manager job applications were made and randomly assigned male or female names. The applications were identical regardless of assigned gender. 127 participants were given these faux applications. After analyzing the “perceived student competence”, suggested salary offer, and how “deserving” the student was of the position assigned to each faux application by the participants, researchers found that both males and females had unequal biases towards the female applicants. Furthermore, “both male and female faculty judged a female [applicant] to be less competent and less worthy of being hired than an identical male student, and also offered her a smaller starting salary and less career mentoring” (Moss-Racusin et al., 2012). The unconscious bias of these participants in this study further supports that biases are a problematic factor that reaches into the hiring process of many corporations. Pattern matching in this case was likely applied on the basis of gender stereotypes. Studies show that in 2020 for every 100 men promoted to managerial positions only 85 women were also promoted. Because of this, women only hold 38% of managerial positions (McKinsey & Company, 2020). Without intentional acknowledgment of this gender disparity, individuals can pattern match males with the ability to be managers due to the greater amount of male managers in the corporate world.
Artificial Intelligence also factors into the modernized hiring experience. AI is often used to cut down on the number of resumes needed to be reviewed by hand. AI in these instances pattern matches employment requirements set the employer to resumes containing this information. Resumes that are not applicable to the AI’s determining factors of success are rejected and removed from the pool of candidates. Amazon, for example, sought to create an AI machine that would take “100 resumes” and “spit out the top 5” for easier hiring processing (BBC, 2018). This AI machine however began to reinforce the current diversity failures of the modern workplace. According to BBC, “The system started to penalize CVs which included the word ‘women’”. This problem was addressed by Amazon and the project was ultimately abandoned (2018). By allowing AI to determine who will be employed, discrimination is ultimately supported. The automatic creation of qualifications given by AI that reach beyond the minimum qualifications for a job removes qualified candidates out of your job candidate pool.
A unicorn company “is a private company with a valuation over $1 billion US dollars (CBInsights). These companies are massively successful and massively impactful for the local economy. These are the companies that investors dream of staking their claim on. Denise Hewett states in her speech that 10.7% of unicorn companies are women-owned, yet only 2-4% of venture capitalism goes to them (Hewett, 2019). Furthermore, Harvard Business Review states that diversity in venture capitalist companies has remained relatively unchanged. “Only 8% of venture capitalist investors are women” and “fewer than 1% are black” (Gompers & Kovvali, 2018). As shown in Denise Hewett’s speech these biased investment choices are a clear form of pattern matching (2018). Investors are encouraged to analyze their own unconscious biases towards social groups and seek to understand if this affects their investment choices.
What are the effects of pattern matching?
Pattern matching causes a lack of diversity in the workplace and reinforces already present inequalities between groups. Consequently, pattern matching has multiple negative effects on businesses.
According to BCG, “increasing the diversity of leadership teams leads to more and better innovation” (Lorenzo et al, 2018). Companies with “below-average diversity scores” have hav average of 26% innovation revenue while companies with “above-average” diversity scores have an average of 45% innovation revenue (Lorenzo et al, 2018). This increase in innovation is provided by the unique outlooks and life experiences of the diverse groups of employees involved. Different social groups have different outlooks, and can then propose different solutions to problems, new ideas, and notice different areas of improvement. Pattern matching decreases this availability of new ideas, by decreasing the amount of diversity in the workplace. This decrease in diversity in turn decreases innovation. One of the key factors of corporate success is innovation ahead of competitors. Without diversity and its innovative properties, corporations have the potential to lose an innovative edge against the competition.
Decreased Financial Performance
The BCG also states that with an increased rate of diversity in leadership teams, “financial performance” improves. BCG relates this to the increased innovation that diversity brings about. Bringing new ideas and “unconventional solutions to problems” from collaborative and diverse teams, causes diverse companies to “outperform their peers financially” (Lorenzo et al, 2018). Since pattern matching decreases the diversity of companies which causes a decrease in innovation, it can be concluded that it also decreases the potential financial performance of corporations. Financial performance is the defining factor of a corporation. By limiting the amount of diversity through pattern matching, financial performance is stunted in its growth undermining the success of a company.
How do we prevent mistakes in pattern matching?
“A diverse mix of voices leads to better discussions, decisions, and outcomes for everyone.” - Sundar Pichai, CEO of Alphabet Inc.
By understanding the previously stated causes of pattern matching (stereotypes and unconscious biases), we can effectively decrease the amount of pattern matching that we use when making business decisions.
Encourage yourself and employees that are part of the hiring and investment processes to acknowledge their own unconscious biases (HARVER). By educating employees on the potential biases that are present and encouraging them to intentionally avoid such biases, a more diverse workforce can be created and pattern matching use can be decreased.
Use a Collaborative Hiring Process
“Diversity attracts diversity” (Gompers & Kovvali, 2018). By increasing the number of individuals and diversity that are a part of the hiring process, more “worldviews” and perspectives can be taken when assessing candidates. Dispersing the responsibility for new hires or investment decisions allows for a decreased risk of bias when the individuals responsible for hiring are a diverse group themselves.
Acknowledge Personally Held Biases
Building more diverse and successful workplaces starts with the individual desire to create an equal playing field for all potential employees. Personally held biases can be those beliefs that are not supported by science. For example, Ali Tamaseb found in his research that some factors that are normally considered to be predictors of success such as “age and university” actually “didn’t matter” when it came to an individual's employability or invest-ability. According to Tamaseb, “What matters is having a history of having generated value, even if that value was small.” (Davis, 2021). Other personally held biases can be those supported by stereotypes (as previously discussed). Some stereotypes are based on gender or race, while others have to do with religion or specific social groups. Regardless, it is important to notice what stereotypes you might unintentionally lean towards and account for this in your decision-making.
What does it all mean?
As we opened this Insights article, for many businesses and industries, we create patterns to reduce errors and standardize production. However this same behavior, when not thoughtfully applied can create unintended negative consequences such as bias. Pattern matching, or the application of stereotypes and unconscious bias to the hiring process, decreases the likelihood of a diverse workplace. This in turn decreases the innovation and financial success of the business. Evidence of pattern matching can be seen in workplace and investment statistics, such as those mentioned by Denise Hewett. By addressing pattern matching and seeking to minimize its effects on hiring and investment, companies can create a more diverse, innovative, and profitable workforce.
“The greater the diversity, the greater the perfection.” -Thomas Berry, CP, PhD, Historian
“Strength lies in differences, not in similarities.” - Stephen Covey, author of Seven Habits of Highly Effective People
Alexandra. (2021, January 26). 11 Ways to Reduce Hiring Bias. Retrieved from https://harver.com/blog/reduce-hiring-bias/#Awareness
BBC. (2018, October 10). Amazon scrapped 'sexist AI' tool. Retrieved from https://www.bbc.com/news/technology-45809919
Bleiweis, R. (2021, March 24). Why Salary History Bans Matter To Securing Equal Pay. Retrieved from https://www.americanprogress.org/issues/women/reports/2021/03/24/497475/salary-history-bans-matter-securing-equal-pay/
Bryant, K. (2019, March 08). Unconscious bias in risk management – a hidden risk. Retrieved from https://www.theprogresspartnership.com/unconscious-bias-in-risk-management-a-hidden-risk/#:~:text=StratgeicRISK spoke to The Progress,relying on unconscious psychological bias.
CBInsights. (n.d.). The Complete List Of Unicorn Companies. Retrieved April 30, 2021, from https://www.cbinsights.com/research-unicorn-companies
Davis, A. (2021, April 25). Discovering the next unicorn founder with data, not gut instinct. Retrieved from https://pitchbook.com/news/articles/unicorns-stereotypes-Ali-Tamaseb-research-founders
Educative. (n.d.). Definition: Pattern matching. Retrieved from https://www.educative.io/edpresso/definition-pattern-matching
Gompers, P., & Kovvali, S. (2018, July 09). Finally, Evidence That Diversity Improves Financial Performance. Retrieved from https://hbr.org/2018/07/the-other-diversity-dividend
Lorenzo, R., Voigt, N., Tsusaka, M., Krentz, M., & Abouzahr, K. (2018, January 23). How Diverse Leadership Teams Boost Innovation. Retrieved from https://www.bcg.com/publications/2018/how-diverse-leadership-teams-boost-innovation
McKinsey and Company. (2021, February 18). Women in the Workplace 2020. Retrieved from https://www.mckinsey.com/featured-insights/diversity-and-inclusion/women-in-the-workplace#:~:text=Not surprisingly, men end up,women hold just 38 percent.
Mcleod, S. (2017). Stereotypes. Retrieved from https://www.simplypsychology.org/katz-braly.html
Moss-Racusin, C. A., Dovidio, J. F., Brescoll, V. L., Graham, M. J., & Handelsman, J. (n.d.). Science faculty’s subtle gender biases favor male students. PNAS, 109(41), 16474-16479. doi:https://doi.org/10.1073/pnas.1211286109
Hewett, D. (2019, November 25). The Problem with Pattern Matching [Video file]. Retrieved from https://youtu.be/wBisUZdAerw
Wiese, J. (2020, December 09). Is Unconscious Bias Coloring Your First Impressions? Retrieved August 27, 2019, from https://emtrain.com/blog/unconscious-bias/coloring-first-impressions/
Copyright © 2021 by Arete Coach LLC. All rights reserved.