Skilled Technical Workforce Classification

• Activity Number: 483 - Privacy and Work Force
• Type: Contributed
• Date/Time: Wednesday, August 10, 2022 : 2:00 PM to 3:50 PM
• Sponsor: Government Statistics Section
• Author(s): Leonel Siwe* and Vicki Lancaster and Cesar Montalvo and Haleigh Tomlin
• Companies: University of Virginia and University of Virginia and University of Virginia and Washington and Lee University
• Keywords: Skilled Technical Workforce; Labor Market Information; Non-negative Matrix Factorization

Abstract

Skilled technical workforce (STW) are occupations that require technical knowledge but no bachelor’s degree for entry. In 2015, Rothwell proposed the metrics for classifying STW occupations using education and knowledge data from the Occupational Information Network Content Model. These data present a number of fitness-for-use issues: small sample sizes, data over a decade old, 196 occupations that have no education and/or knowledge data; plus these metrics are calculated ignoring the standard errors of the estimates. STW is a function of the nature of work which is rapidly changing due to emerging technologies. The data used to quantify these metrics needs to be current. We propose a new approach for classifying STW occupations. We use labor market information which provides detailed information at the occupation level, is updated daily, contains O*NET-SOC codes, minimum education requirement, and detailed information on technical skills. As a proxy for knowledge, we evaluate the use of skills listed in these job-ads. In lieu of a discrete criterion for knowledge we use the machine learning method, Non-negative Matrix Factorization, to separate occupations into STW and nonSTW.