Interpreting Longitudinal Data, Part 1 – Looking to the Past

Let me start this chapter with a piece of Chinese wisdom, not because of its superficial similarity to the ‘mountains’ in the graph of a time-series, but because it expresses the evolution of my own thinking on this matter. The present state of time-series analysis seems captive in the second phase mentioned by Chin-Yuan Wei-Hsin, where things as vital as waters and mountains, under the guidance of learned teachers, are understood to be something else. In analogy, statistical concepts and ideas that have ignored the true nature of socio-economic data dominate our approach to their longitudinal analysis. Chapters 2, 3 and 4 laid the foundation for a more realistic approach to time-series, arriving finally, like the old monk, at the unromantic realization about the true nature of ‘mountains and waters’.

Thirty years ago, before old monk had studied Zen, he saw the mountains as mountains, waters as waters. Later when he came to know a good master and was first initiated into Zen, he no longer saw mountains as mountains or waters as waters. Now that he had got a resting place, he again sees that mountains are only mountains and waters only waters ¹