Emerald Management Reviews

Design/methodology/approach - Argues prior research suffers from not distinguishing restatements properly. Develops three criteria for distinguishing errors from irregularities: (1) by detecting the use of 'fraud' or 'irregularity'; (2) by identifying linked official investigations of irregularity and (3) by identifying the presence of other independent investigations. Applies to all filings 2002-2005, and adds validation tests. Then links to turnover of chief executives (CEOs) or chief finance officers (CFOs), and extends to 1997-98.

Findings - Finds that restatements due to irregularities comprise 24% in 2002-2005, but 48% in 1997-1998, and exceed in number those identified by prior studies. Shows that the turnover of CEOs/CFOs has remained in line with the incidence of irregularities; and that the frequency of restatements due to errors has inceased.

Research limitations/implications - Urges researchers to adopt this method in order to increase the power of their tests for accounting irregularities.

Originality/value - Presents an improved way of separating villains from the careless.