The study compared two models, GJR-GARCH and EGARCH, to predict stock market volatility. These models account for the asymmetry between positive and negative errors in volatility. Data from S&P 500, NASDAQ Composite, and NYSE ARCA Oil and Gas Index were used from 1968 to 2002. The researchers found that the EGARCH model was the most accurate in predicting volatility for all three datasets.