A vibrator was operated at different pressure (psi) levels and the impact of the vibrator (Hertz) was measured at different distances away from the vibrator. Measurements were taken 1-hour daily for 21 days. This generated more than 300 pairs of pressure (psi) and vibration (Hertz) datasets.
The objective is simple:
- To estimate the effect of pressure (different pressure levels applied) on the observed vibration impact (measured in Hertz) at the different distances away from the vibrator.
- The thought is that regression analysis can suffice to measure the relationship between pressure and vibration impact, but the question is, given the independence assumption of regression analysis, most often violated by time series data, what can be done in this situation. Note dataset for some points (locations) are more than 300 pairs, others are simply about 20.
- Please, advise. thanks.
I would say yes. In this case you are probably looking for a correlation of source properties, amplitudes and distance. I would call this more of a measurement than a statistical analysis.
Statistical methodes and different types of regression is used regularly in such cases. My own tests of bending wave losses from a point source on a concrete slab has beed done using a logarithmic fitting of the data giving a faster decay the first meters and a slower decay over longer distances.
- Badges
- Science topic
- Similar topics
- Econ
- Innovation Economics
- Innovation