Most Recruiters Not Fully Confident in Applicant Screening Methods

 

Survey finds talent professionals would like to use more-objective selection tools.

Talent acquisition professionals fear that they are losing out on qualified job candidates by relying on traditional applicant screening methods.

Only 20 percent of 520 recruiting professionals said they were fully confident in their employers' overall ability to effectively assess the skills of entry-level applicants using time-honored selection methods like interviews and application and resume scans, according to a survey conducted by global consultancy Mercer and the Society for Human Resource Management (SHRM).

Most employers use in-person interviews (95 percent), application reviews (87 percent) and resume reviews (86 percent), but nearly one-half of respondents said they have "little or no confidence" in application and resume reviews.

"Since application and resume reviews are typically the first line of screening for job applicants, many candidates never even get to the interview," said Barb Marder, a senior partner at global consultancy Mercer. Respondents expressed much more confidence in using in-person interviews to assess candidates. Marder added that entry-level applicants without any work experience often have trouble getting past the review phase because HR dismisses them for lack of experience.

Not knowing enough about the applicant was the most commonly reported concern in assessing candidates during the early phase of the recruitment process, the application review. "Job applications have been scaled back to not much more than contact information, job title and tenure," said Joe Murphy, an executive vice president at Shaker, an assessment software company based in Cleveland. "They pretty much offer no insight to job fit."

Murphy added that resumes "are plagued by several underlying limitations" such as a lack of structure and candidates' attempts to tailor content to meet job descriptions and keyword-matching capabilities in applicant tracking systems.

Due to the unstructured and random nature of resumes, it is difficult to attribute measurable value to them, he said. "Resumes [and applications] are unsuitable for predictive analysis, and career experience as listed on a resume is a poor predictor of job performance."

Pre-Employment Assessments: A Better Selection Method?

Less than half of companies use selection tests for entry-level hiring (42 percent), the survey showed. Few organizations use personality tests (13 percent), cognitive ability tests (10 percent) or online simulations (2 percent) to select entry-level employees, methods which some research suggests could be more accurate predictors of performance.

Murphy believes assessments are superior to resumes and job applications as a selection method due to such tests being:

  • Intentionally designed and assembled to capture specific information.
  • Standardized, methodical and scorable. "Assessments by their nature comprise responses that lead themselves to adding numeric value, allowing for ranking," he said.
  • Conducive to predictive analytics.

Not all assessments are predictive, Murphy said. He explained that conducting an in-house validation analysis with an assessment is the critical step required to establish a statistical relationship between the assessment and on-the-job performance. "Locally validated assessments transform the practice of sorting resumes into an objective top-down ranking of candidates from more capable to less so," he said. "Using this approach, recruiters can quickly identify and invest time with the most qualified applicants."

Strong Future for Assessments

Machine learning and artificial intelligence (AI) hold intriguing promise to improve pre-employment assessment of candidates in two major ways, said Ji-A Min, head data scientist at Toronto-based Ideal, which builds recruitment automation software that uses AI.

"Machine learning algorithms can be applied to nonconventional datasets such as video and phone interviews in order to extract insights on candidates' personality, skills and cognitive ability, while AI can learn to combine data sources such as resumes, social profiles, assessments and interviews in novel ways to analyze how qualified a candidate is for a role."

By Roy Maurer

Source: shrm.org