Detrended fluctuation analysis for postural control data

The implementation of the non-linear analysis method, Detrended Fluctuation Analysis (DFA), has been hindered by a lack of consistency in its application. Our study sought to establish recommendations for the box size ranges over which the output of DFA, alpha (α), is calculated for postural control data. It was found that using a minimum box size of 64 and a maximum box size of 128 was ideal for postural control data. The resulting α value, however, is significantly different from what would be obtained by using, other, more generic recommendations, suggesting the need to increase standardization in the field and pay more attention while comparing results across studies.