Hiring? It's an AI vs Psychometric Assessments Showdown

Artificial Intelligence (AI) is a powerful tool in the hiring process. With predictive analytics and natural language processing (NLP), AI systems can sort through and filter countless job applications with unprecedented speed. However, the current state of AI systems in hiring and its potential for the introduction of bias may lead to the exclusion of candidates that don't conform to a predictive norm.

If you are trying to find the right person for the job — and you should be — psychometric assessments can be the most reliable and predictive tool available.

Diverse in their scope and accuracy, psychometric assessments in the broadest terms seek to measure things like behavior, cognition, emotional intelligence and psychological or psychometric fit. During the hiring process, HR professionals and hiring managers use psychometric assessments to gain insights into candidate aptitude, reasoning, skills, motivation and overall personality.

There are many big names in the psychometric assessment space, including Myers-Briggs™, The DiSC™, Strength Finder 2.0™, etc. The most accurate and reliable psychometric assessment is a relative newcomer to the market, the Core Values Index. The CVI reveals the unchanging, hardwired psychometric nature of the individual with 97.7% repeat reliability — take the CVI now and again in five years and your scores are likely to vary by less than 3%.

In the talent acquisition space, psychometric assessments like the CVI don't measure experience or professional qualifications, they are used to determine a candidate's psychometric fit for a role.

Accuracy

When it comes to hiring, recruiters want the most accurate information to help them determine candidate fit as efficiently and quickly as possible. AI helps fill this need. Automated systems scan candidate resumes looking for keywords¹ that match the job listing. If a position requires two years of data entry skills but a resume doesn't have those words, the system can reject it automatically. Advancements in heuristics are making this process more accurate every day.

Experience isn't the only indicator of skills and motivation, though. Like a laser guided targeting system, the CVI psychometric assessment increases search accuracy by identifying candidates that have the personality DNA most appropriate for success in the role.

The Core Values Index works by asking the candidate to pick two words out of groups of four, and then matching personality traits like leadership, intuition, problem solving, and knowledge gathering to those selections. It sounds simple but it yields results that are surprisingly accurate and extremely difficult to skew. It is nearly impossible for the candidate to skew their answers to yield a particular assessment outcome.

While AI can select resumes that match the qualifications and hard requirements of the position, a CVI assessment provides a more overarching perspective on how well a candidate is likely to succeed in a specific role. Thus, psychometric assessments outpace AI for accuracy.

Bias

AI systems can be biased. Despite their automated nature, they still require programming of some sort. During this process, the people that train the AI system can unconsciously or unknowingly instill their own prejudices as well as biases that may be inherent in the data. The predictive algorithms in the hiring process would then hold those same preconceived notions.

For instance, research found that² predictive algorithms have shown concerning racial biases that treat Black individuals differently and negatively. This, in part, came about because the data used to train the AI systems contained information about predominantly white candidates. The same concept can happen if the AI system is biased toward men, seeking similar candidates for a specific role.

Psychometric assessments like the CVI that use a non-skewable system are bias-free, however. They ask candidates the same questions without taking into account any background or demographic information of the individual. While HR professionals and others in the hiring process can have their own unconscious biases, a CVI assessment is typically taken before the interview phase. CVI profiles can be used to narrow down the list to a select group based solely on psychometric fit before any profile information open to bias like age, gender or ethnicity is revealed or considered.

Using the CVI psychometric assessment, the hiring process can proceed with a greatly reduced risk of unfair or biased consideration of candidates. Employers can also avoid potential liability over discrimination.

Fact Over Fiction

Psychometric assessments can struggle against a mixed reputation. While Core Values Index assessments are accurate and based on repeatable facts, other big name psychometric assessments are not. For instance, the Myers-Briggs Type Indicator (MBTI) — arguably the biggest name in the industry — is widely used but is not based on any concrete science or evidence. In fact, much of the MBTI's foundation has been roundly discredited and has been determined to be about as accurate as a coin toss. (The MBTI struggles to reach a 70% repeat reliability rating in the best circumstances, more often averaging around 50%³.)

In this case, AI would make for a better resource for hiring. However, the Core Values Index assessment is the best option since it identifies the individual's unchanging psychometric hardwiring. It is highly appropriate not only in the workplace, but for personal use as well.

In the near future, AI recruiting systems may integrate with psychometric assessments like the Core Values Index. This combination would optimize the recruitment process, improving the accuracy of viable candidates that go on to the interviewing stage.

AI vs. CVI

When compared to most psychometric assessments, AI may be the winner. However, the Core Values Index provides more accuracy, reduces or eliminates bias, and uses repeatable and accurate psychometric data to find the best fit for a position.

---

NOTES:

1 https://www.vox.com/recode/2019/12/12/20993665/artificial-intelligence-ai-job-screen

2 https://arcadia.io/arcadia-statement-on-predictive-analytics-and-bias/

3 https://jobtalk.indiana.edu/HRMWebsite/hrm/articles/develop/mbti.pdf