Surveys Show Online Employee Reviews May Predict Business Fraud


Anonymous reviews of employees leaving online could be used to anticipate and thwart corporate fraud, new research finds.

According to a study by researchers at Harvard Business School and Tilburg University in the Netherlands, information extracted from employee reviews left on the company review site Glassdoor.com can be easily used to other things such as the performance of the company, the press and the press. Helps predict fraud beyond observable factors. Industry Risks and Previous Violations.

The review provides employee insight into company management practices, culture, operations and performance pressures that can contribute to fraud risk, said Dennis Campbell, professor of business administration at Harvard Business School. to say. Tilburg School of Economics and Business Administration. Hearing that “background tone” provides an early warning of potential fraud, he says.

“Our theory is that the reason people cheat is actually the environment they find themselves in,” says Dr Shan.

Anonymous Notice

For the survey, the researchers extracted anonymous review information from US companies listed on the employee review site Glassdoor.com from June 2008 to December 2016. Companies with less than 10 reviews in the period were excluded.

Then, from 2008 to 2017, we acquired data such as company size, capital structure, profitability, and topical data such as the number of press articles related to each company. A business that does not have the necessary variables and data, such as a business that consolidates all the data and goes bankrupt or is acquired. The final sample consisted of 13,363 observations from approximately 1,478 companies.

Finally, we extracted the 26,934 malpractices committed by US public companies between 2008 and 2017 from Violation Tracker, a search engine for civil and criminal cases filed against companies. .. This allowed us to identify disproportionate comments in the opinions of companies found guilty of illegal activities.

Using machine learning technology to understand how many “cheat words” such as bureaucracy, compliance, discouragement, favor, harassment, hostility, meritocracy and thoroughness are included in the assessment of ‘a company. , Creation of a risk measure to predict future harassment violations. , say the researchers.

Value and limits

Hui Chen, a former Justice Department compliance consultant, says that while this type of analysis is valuable, it is important to be aware of the limitations of the investigation. The researchers measured illegal activities based on government sanctions, so their methodological predictions are “hidden illegal activities” that are unknown or not pursued by the government for various reasons. She says she may miss a bit.

Dr Campbell says that although risk indicators have been created and validated in the fraud cases observed, they believe they can be used to identify potential cases of “hidden” fraud.

Maxy is a writer in Union City, New Jersey. [email protected]..

Copyright © 2021 DowJones & Company, Inc. all rights reserved. 87990cbe856818d5eddac44c7b1cdeb8


Source link

Previous Phoenix Interviews Candidates For Director For Office Of Accountability And Transparency
Next Three city polls show varying opposition to Vancouver's mandatory parking permit

No Comment

Leave a reply

Your email address will not be published. Required fields are marked *