The article discusses how observations in a time series can show a pattern of correlation at fixed time intervals, known as autocorrelation. This autocorrelation function is important in understanding time-series data. The researchers define and explain how to estimate the autocorrelation function and partial autocorrelation function.