Quality has no single agreed definition, which is both its flexibility and its weakness. Common inputs are return on equity, earnings variability, accruals, debt to equity and gross profitability. Two quality indices can hold materially different portfolios.
The appeal is defensive: quality screens have historically reduced drawdowns in equity sell-offs, because companies with strong balance sheets are less exposed to funding stress. The cost is that quality often overlaps with expensive, which can cancel a value-factor tilt held alongside it.
When a definition is loose, test it. Check whether the screen's historical result survives being computed with a different but equally reasonable metric; premia that only exist under one specification deserve scepticism.
Related: factor-investing, profitability-factor, value-factor, low-volatility-factor, fama-french-three-factor, eps