In this paper, we investigate different characteristics of the correlation in the time series of the difference and the volatility by using the passive and active measurements of the data network traffic. In particular, we try to understand the long-range correlation in the volatility in terms of elaborate analysis. To this end, we expressed the difference and the volatility in terms of positive and negative variation, and applied the detrended fluctuation analysis to all quantities in order to investigate the long-range correlation. We found that both the volatility and the positive and negative variations showed long-range correlations, while the difference exhibited an anti-correlation. In the case of the difference, due to an asynchronous fluctuation between the positive and negative variations, the fluctuation is cancelled and exhibits an anti-correlation. In the case of the volatility, in contrast, the variations fluctuate synchronously which results in the amplification of the volatility fluctuation. As a result, the volatility shows a long-range correlation. We also applied the various time delays to the difference and the volatility, and found that the results are almost independent of the time delay.