When a stock looks cheap on earnings, what evidence would make you doubt the earnings themselves?

A low price-to-earnings ratio is often the first thing people point to when a stock looks like a bargain. But the denominator is an accounting figure, and accounting figures can be shaped by timing, one-off gains, capitalised costs and changes in share count. A stock can look cheap on trailing earnings while free cash flow tells a more modest story, or the reverse. The devil's-advocate version of the question is simple: what would you need to see before you stopped trusting the headline number? Some people compare earnings with operating cash flow over several years, check whether receivables are growing faster than sales, or look at how often "adjusted" figures diverge from reported ones. Others ask whether the cheapness persists after normalising margins to a full cycle rather than a peak year. None of this proves a stock is overpriced. It just makes the cheapness claim testable. Which check do you find most useful in practice, and what result from it would actually change your mind?