Detrended Fluctuation Analysis (DFA) is a well-known method to evaluate scaling indices of time series, categorizing the dynamics of complex systems. In the literature, DFA has been used to study the fluctuations of reaction time RT(n) time series, where n is the trial number. Herein we propose treating each RT(n) as a duration time that changes the representation from operational (trial) time n to chronological (temporal) time t, or RT(t). To do this, we fill each time interval of RT(n) with fixed noise of magnitude 1 and with a randomly determined sign. The fixed noise represents the rigidity (order) and the random change of sign represents the flexibility (randomness) of the generated time series RT(t). Then the DFA algorithm was applied to RT(t) time series to evaluate scaling indices. We show that this new perspective leads to better results in: 1) differentiating scaling indices between low vs. high time-stress conditions and 2) predicting task performance outcomes. The dataset we analyzed is based on a Go-NoGo shooting task which was performed by 30 participants under low and high time-stress conditions in each of six repeated sessions over a three week period.