Design Of Hiring Algorithms Can Double Diversity In Firms

from Fast Company

We know that algorithms can outperform humans across an expanding range of settings, from medical diagnosis and image recognition to crime prediction. However, an ongoing concern is the potential for automated approaches to codify existing human biases to the detriment of candidates from underrepresented groups.

For example, hiring algorithms use information on workers they have previously hired in order to predict which job applicants they should now select. In many cases, relying on algorithms that predict future success based on past success will lead firms to favor applicants from groups that have traditionally been successful.

But this approach only works well if the world is static and we already have all the data we need. In practice, this simply is not the case. Women, for instance, have been entering STEM fields in record numbers, but if firms used their historical employment data to decide whom to hire, they would have very few examples of successful female scientists and engineers. At the same time, the qualities that predicted success yesterday may not continue to apply today: just think of how remote work during the pandemic has changed the nature of teamwork, communication, and teaching.

So instead of designing algorithms that view hiring as a static prediction problem, what if we designed algorithms that view the challenge of finding the best job applicants as a continual learning process? What if an algorithm actively seeks out applicants it knows less about, in order to continuously improve our understanding of which candidates will be a good fit?

More here.

Posted in Business, Careers, Law, Technology and tagged , , , .

13 Comments

  1. Li’s article was very interesting, and I think using algorithms to find the best possible applicant for the job could be a solution for certain groups of people to get the jobs they deserve. Simply looking at the NFL, most team owners of NFL teams are white and continuously higher white coaches because that is what they have been successful with in the past, even if there is a minority candidate who is better suited for the job. Like the article pointed out, it could be a gateway for women who are in STEM, which has been a field mostly dominated by men, to get the recognition they deserve. I do like how the article pointed out how it would be inefficient to hire people based on their business’ previous success because times are always changing. COVID 19 has completely changed the dynamic of working. Companies now use virtual platforms, and it has worked quite well. Furthermore, with these better applicants being discovered because of this algorithm, it will certainly lead to more business production, especially if the algorithm updates over time. Using the same hiring algorithm does not make sense because of how fast the world changes. It should not matter on the race or gender of applicants. Everyone deserves an equal chance. If someone is the best candidate for the job, they should be holding the job. If this gets implemented, it will increase the amount of opportunities minority’s get in the job field. One thing that I do not think will ever be able to get replaced in the hiring process is the face to face interview. I do believe this algorithm could be used to decide who gets the chance at an interview, but I believe the interview needs to be held in a face to face atmosphere. Personally, I think I would talk differently to a computer as opposed to when I am talking to a person. The face to face aspect of the interview allows the interviewer to see an aspect of the candidate that I do not think an algorithm can capture. It allows the interviewer to see the type of person they are and if they are able to handle the daily stresses of conversing in the job world. All in all, I think algorithms certainly have a place in the workforce especially for diversification and finding the best possible candidates regardless of race or gender.

  2. The year 2020 has taught us a lot, one of the most important things being diversity. Diversity is the things about us that make us different. It is not a new concept, but just recently do I think that people are starting to realize its importance. Diversity is very important anywhere, but I think it is very important for a business to have diversity amongst their community. Some of the benefits of diversity are Cultural sensitivity, insight, and local knowledge means higher quality, targeted marketing, also diverse teams are more productive and perform better, and local market knowledge and insight makes a business more competitive and profitable, lastly, a diverse skills base allows an organization to offer a broader and more adaptable range of products and services. There are many more ways than it is beneficial; this is just scraping the surface on how diversity can benefit a workplace. I think that it is important to have this algorithm if its main purpose is to produce diversity in the workplace. There is a demographic when we think of these people who oversee most companies; white and middle-aged. With this algorithm, we can break that up and have more diverse people working in different positions. I like how they were able to test this algorithm out and give us real feedback. I also think were some very disheartening statics that explains the discrepancies and flaws with the hiring process. The lab test revealed some of the problems and how the algorithm helps. I think that this algorithm and other programs like it are important because they promote diversity in the workplace. Everyone should have a fair and equal opportunity for success.

  3. Algorithms are used constantly in every day life, a common one being social media. If you search the hashtag jeans, ads will begin to pop up in your feed pertaining to buying a new pair of jeans. Hiring algorithms can be the next step in perfecting calculations. What the algorithms in the workplace would do it take the information on previously hired and successful employees and use that data in their next hiring process. The problem with static hiring is that there is currently not much diversity and if you take the data where there is not a diverse work environment, you are still hiring with a selective group. As of now, there is a great need for diversity in the workplace and inventing a new algorithm might just be the answer we are looking for. Raymond and Bergman developed a new algorithm where it values exploration to learn why people had not been considered for other jobs. This is the first step in a great direction where we can develop a proper algorithm that can accurately chose employees for the hiring process that is not biased or discriminatory. When you use exploration in algorithms, you can hire individuals with more quality and content. By using algorithms, the hiring process can be greatly improved and that bias , whether you want it to or not, that was once there can no longer affect the reason for either getting the job or not. These are the first few steps in a new and improved direction where the hiring process can have justice and reasoning for the people that are being hired and continuing in this direction, the workplace will be a changing and growing place.

  4. Using programs to hire people can help companies to hire people quicker and to see who is the most qualified. Personally I think it’s hard to hire people online because you don’t actually get to meet the person. I would like to see how people interact with their bosses or other people in the office. It is always important to meet people in person before you hire them. Hiring someone just based off of their resume or past trends with other employees may be efficient, but you don’t actually get to meet that person. Personal relationships are very important in the workplace and if someone is hired then they end up being a terrible person or if they don’t interact well with people in the workplace then you are stuck in a bad situation. If I were hiring someone then I would like to get to meet them in person and see their personality and other things on top of their resume. You can’t get to know someone based off of a piece of paper, you get to know them by meeting them face to face. Nowadays that would more likely be over Zoom, but it is still better than contacting someone over email and hiring them based off of their resume. These algorithms are good to find potential candidates for jobs. You can use the technology to look at their resumes and other pieces of information that could make people stand out, but the interview would really be where the companies decide if they want to hire someone. Overall I think that hiring programs can be helpful in finding good candidates for a job, but I think the interview in person or online is extremely important to do before hiring anyone.

  5. Hiring is rarely a single decision, but rather the culmination of a series of smaller, sequential decisions. Algorithms can play different roles throughout this process: Some steer job ads toward certain candidates, while others flag passive candidates for recruitment. Predictive tools parse and score resumes, and help to hire managers to assess candidate competencies in new ways, using both traditional and novel data. Therefore algorithms can certainly improve efficiency in the hiring process, but it can also bias due to the lack of data. The article addressed this issue by saying that a suitable algorithm would “view the challenge of finding the best job applicants as a continual learning process” and “actively seek out candidates that it knows less about.” However, crucial traits of a prospective candidate such as problem-solving skills and critical thinking can’t be detected through patterns. For instance, the essence of critical thinking is being able to think outside the box and think beyond patterns, job applicants are tested to see if they have the capability to apply that to their jobs. Past successes that can be quantified and recognized by algorithms do not demonstrate the qualities of critical thinking, even if it were to be able to recognize critical thinking, it won’t be able to predict how the candidates will apply it for their jobs. Additionally, algorithms can bypass the possibility of future growth and the potential growth of a job applicant can prove to be more valuable than a static one. The unpredictability of growth can jeopardize an algorithm’s ability to find the best applicant, and if the algorithm has the same issue that a normal hiring process has, there is no reason for companies to invest in another hiring method. Perhaps a combination of an algorithm and hiring team would be best because solely depending on an algorithm and assuming that it only betters in selecting the best applicant would eventually bring unfairness to the hiring process.

  6. Applicant tracking systems (ATS) is a software that is used by companies to help sort and qualify all the applicants that are actively applying for positions. The purpose of the applicant tracking system is to collect all of the candidate resumes that are received for a position, and then translates that into a basic format that then uses these standardized resumes to compare them to the job description and then selects the one that is the best fit for the function of the job. So, when the recruiter must decide which candidate to select for a particular position, they will automatically know who the ideal candidate that best fill the job position. Study shows that the (ATS) on the average will only approve about 25% of its initial resume applications, this happens because the automated system scans a large quantity of applicant resumes but rejects a substantial amount of them. This is because the hiring manager defines the necessary skills and key words that needs to be included in the resume for what they believe will best fill the functions of the job. The approval rate of candidate’s resume is contingent upon the previously hired candidates so this way the hiring manager will be more susceptible to hire someone that has similar job skills, and requirements as the previously hired. So to explain succinctly, when the firm utilizes algorithms they will select candidates that have been proven to be a success based on previous successors resulting in a undiversified workforce, so the underlying issue here is that many firms will consist of a workforce that will lack diversity.

    My Opinion

    I believe that if firms were to utilize automated approaches that codify previous candidates biases to the selection of underrepresented candidates, I think that this different form of applicant tracking system will have the potential to serve the workforce and firms to a greater degree. The basis of this evaluation is to consider that candidates who belong to the minority groups such as Black or Latinos will have the potential to obtain a job offer. Additionally, this can also increase the various types of skills and qualities that a corporate structure consists of resulting in a structure that is diverse and has an array of different attributes.

  7. This article is very applicable to the technology run world we live in today. There are so many different variables when it comes to algorithmic selection of job applicants. The article says though, “Women, for instance, have been entering STEM fields in record numbers, but if firms used their historical employment data to decide whom to hire, they would have very few examples of successful female scientists and engineers.” This means that we should use algorithms in combination with the new standards of who works in what job in today’s time. The fact that they are using multiple algorithms to make sure they can have a balance in the ways of diversity and gender selection of workers. We are subjected to many algorithms in todays time, some on the internet determine what ads we see and what videos pop up on our feed. When I saw this statement, “This suggests that the additional Black and Hispanic candidates the algorithm selected were just as good as other candidates—the firm had simply not given them as many opportunities in the past. We found a difference in gender results, too. All of the algorithms increased the share of selected applicants who are women…. 50% with the updating SL model.” This shows that algorithms don’t have any serious bias when it comes to selection. Many companies run by one race/gender as the majority like to keep their new employees that are entering their business the same race/gender as themselves. But by using algorithms we can achieve more diverse outcomes when it pertains to hiring employees. With the lack of in person interviews/ in person anything for that matter, we can use these online resources to our advantage. In my opinion, striving to achieve more diverse employment numbers are great because it introduces new backgrounds to the workplace. In turn, opposing viewpoints and ways to think about things differently. What provoked my opinion was this statement, “When you incorporate exploration into the algorithm, you improve the quality of talent and hire more diverse candidates. Firms that continue to use static approaches in their algorithms risk missing out on quality applicants from different backgrounds.” This shows that using these types of algorithms should be the hiring method used from now on as I can’t see a reason why a company would advocate against diversity.

  8. Choosing the right person for a job can be challenging. The sheer number of resumes can be overwhelming. It may be very hard on a boss to hire someone. if its a big company they might need experienced people and show people how its done here, and they need them to be on top of their things. As small companies can take a new one and it will be a ” you have to learn your way around here” and they’re okay with that. To see the close relationship between algorithms and hiring, consider the simple fact that hiring is essentially a prediction problem. Some may have wrong predictions and some may have right ones, you just kind of got to go with your gut, and hope for the best, and make sure he or she is on point with all the work that is due. When a manager reads through resumes of job applicants, she is implicitly trying to predict which applicants will perform well in the job and which won’t. This can be very stressful because if the boss picks the wrong worker it will look bad on them and that is not a good look for the job or company. the company is looking to increase its salary and one way to do that is to make sure your employers are all doing the correct things and making sure they take care of their responsibilities. If a worker is showing up late and not doing his or her job, they make double check if they picked the right person, and some of these jobs don’t care whether ho or what someone said about you and if you don’t straighten up your work you may be fired before you know it. ” its not about all the good things you do, but if you mess up once, it may be over with you”. that is just how life works and its unfair but what can you do. Managers are looking for consistent hard workers and trying to improve their business and if you aren’t doing that, you are not the right fit. Especially with all this technology now, they can track you in what you do and where you are, so it can be very stressful but its just the way life works.

  9. Programs being utilized to hire people do offer a chance for companies to be more efficient and specific in their efforts. I do find it somewhat flawed however, as the in person experience of the hiring process is essential to a proper hiring system. People are much much different online than in person, and in person their true personality tends to shine. Online they can come off as a completely different person with different actions and different attitudes, so getting an idea of how they are in person is essential. An interview should be required regardless, the system can find likely candidates but to further establish a professional relationship an interview should be needed

  10. From what I understood from this article, these new hiring algorithms seem pretty great. Because what they mean is that women, men, and people of different ethnic backgrounds will be selected on their successes. Versus them being discriminated during the hiring process. Or people being hired for reasons outside of their actual performance. If this understanding was correct it means that the strong will survive and it will definitely increase the competition. The competition would most likely be striving to out perform the others. The last paragraph was interesting as well and goes along with this new algorithm development. This is because that new algorithm would be constantly trying to diversify and try and find different people to fill in the gap to figure out what kind of people would be best suited for a job. My opinion on this is that it definitely is very cool to think about and I think in a legal perspective it would decrease the number of lawsuits based on discrimination. However, it would also have to prove to be worth the investment and should bring those companies profit for which it is being used. This is because if it takes too long to find the ideal people for a specific task, the company could be losing money instead of earning profit.

  11. This article does well to illustrate a new dilemma developing alongside the increasing utilization of algorithms. Algorithmic bias may be inherent in many common algorithms used across society. While algorithms are getting more and more sophisticated and allow for increased efficiency, the existence of such biases may have notable societal effects, particularly in regards to diversity. As the article notes, algorithms are generally based on date that revolves around prior successes within an organization. While this may be effective in regards to continuing to hire people with highly similar qualities to successive employees, diversity is placed at great risk here. Minority groups or people who may come from colleges or majors not favored by algorithms may be systematically excluded from hiring processes and the like, despite potentially being just as valuable of an employee as candidates that may be deemed more favorable by existing algorithms. The fact that many organizations are utilizing automatic résumé screening processes to some degree makes matters worse for such groups. With such a heavily reliance of algorithms now automatically disregarding potentially valuable candidates with no questions asked, it is clear that regard must be paid to algorithmic bias.

    The study conducted by Danielle Li, Lindsey Raymond, and Peter Bergman seems to have successfully expressed a solution for handling such algorithmic bias. They had created algorithms of their own that would be used to analyze the hiring process of various job candidates. Key to one of the algorithm’s design was a focus on exploration in regards to candidates, essentially being trained to seek out candidates with qualities not necessarily present in the more consistently favored candidates. This algorithm, dubbed the UCB model, study seemed to be a success, resulting in more candidates belonging to minority groups that possess valuable qualities being favored while the algorithms. What stuck out to me the most here was that the group did not train the algorithm around any specific diversity factors. More diverse groups of people were thus naturally found by the more exploration-based UCB model, rather than being the direct result of simply training the algorithm to favor these results. This trend was consistent throughout the use of the UCB model, ultimately indicating that diversity could be increased through the use of more exploration-based models. I feel as though this concept in algorithmic design should be applied to more and more algorithms where possible, as it seems to be effective in mitigating algorithmic bias while still ensuring its results are accurate.

  12. Diversity is the presence of differences within a given setting. In the workplace, that can mean differences in race, ethnicity, gender, gender identity, sexual orientation, age and socioeconomic class. Equity is the act of ensuring that processes and programs are impartial, fair and provide equal possible outcomes for every individual. Inclusion is the practice of ensuring that people feel a sense of belonging in the workplace. This means that every employee feels comfortable and supported by the organization when it comes to being their authentic selves. As a woman who is Muslim and black in America, diversity, equity, and inclusion are at the forefront of my mind when seeking potential employers throughout my professional journey. I was born and raised in Queens, NY, the most ethnically diverse area in the world, so I have always interacted with people who are different from me. After moving to Lyndhurst, NJ, a predominantly white township, I realized how important it is to feel represented and welcomed. I desire to work with a company that values diversity so that I can continue to connect with different perspectives while being transparent.

    The COVID-19 pandemic has changed the workplace and lenders have had to shift to more intentional practices related to diversity, equity and inclusion (DEI). To intentionally and sustainably address DEI, organizations need to reassess and integrate their talent management and DEI priorities. Human resources (HR), employee resource groups (ERGs), DEI teams, and the business will need to work closely to accomplish this. Some efforts include but are not limited to: fostering a culture where it’s okay for employees to speak up; putting the company’s diversity, equity, and inclusion mission statement in writing; analyzing the strategies the company’s talent acquisition team uses to attract new employees, etc.

    The study conducted in the article found that when companies incorporate exploration into a hiring algorithm, they improve the quality of talent and hire more diverse candidates. I believe this new approach can give businesses a hand in improving their DEI. As stated in the article, firms that continue to use static approaches in their algorithms risk missing out on quality applicants from different backgrounds. When companies include members with diverse backgrounds and experiences, they are better able to recognize the needs and interests of different stakeholder groups. Machine learning can be used in a wide variety of other tasks such as college admissions decisions, credit approvals, fraud detection, search result optimization, and dating site matching. This algorithm can also be used further, outside of hiring to improve DEI.

  13. Large firms all throughout the world face countless obstacles during the hiring process. While all of such are important, only certain issues relate to the presiding article. Within the article, “Design Of Hiring Algorithms Can Double Diversity In Firms”, a common theme of diversity is followed closely, with an urgency for firms to push their hiring process into becoming more and more diverse, in essence creating an overall more diverse workplace. One company in particular, Fast Company, a research firm which seeks out to research the common workplace, experimented with a specially designed algorithm which seeks through countless applications, attempting to create the most diverse work environment in many different aspects. These include but are not limited to race, nationality, ethnicity, religion, area of residence, major, place of education, level of education, place of birth, family size, etc. The self-learning algorithm teaches itself to adapt to changes in real-world demographics and how they can help or hurt the workplace. In its conclusion the article explains how Fast Company had successfully created an algorithm that can create a more diverse environment.
    In my personal opinion, I believe that there are steep downfalls along with so-called benefits that people may take away from this diversifying algorithm. To begin with, I believe that in its entirety, race should not be a factor in the hiring process. The purpose of this algorithm is to create a work environment that is diverse. However, this process is, fundamentally, leading to a means of hiring that considers race more than it would without the algorithm. To my understanding, whole idea of racism is founding through being conscience of it and making decisions based off of it. This sounds awfully like that definition. With this being said, even though this process may seem progressive and beneficial, fundamentally it makes us more conscience of the important of race than it really should be. Through my eyes, race means nothing. Someone’s race doesn’t make them good nor bad. I see everyone as equal until they prove me otherwise. Therefore, I believe that race shouldn’t be considered whatsoever in anything let alone the hiring process. People who look to see how diverse a company is and therefore make judgement upon it in its entirety because of it, need to second guess themselves. People are people, no matter the color of their skin, nor their religion, nor their ethnic backgrounds. This is what I believe to be true and this is how I would personally monitor my own hiring process.

Leave a Reply to justin grotyohann Cancel reply

Your email address will not be published.