GLD223
- 14
- 7
The discussion confirms that the observed pattern resembling a Gaussian distribution is coincidental, attributed to moisture wicking through render, leading to mold or algae growth. The conversation highlights that many natural relationships between variables can appear linear, quadratic, exponential, sinusoidal, or Gaussian depending on the scaling of axes. It emphasizes the importance of initial curve fitting in scientific analysis, while also noting that the observed data does not meet the criteria for a true Gaussian distribution due to its pointedness.
PREREQUISITESThis discussion is beneficial for statisticians, data analysts, building engineers, and anyone interested in understanding the nuances of data representation and moisture effects in construction.
Nearly every natural relationship between variables (within some arbitrary range) either looks linear, quadratic, exponential, sinusoidal or gaussian. Change the scales of the x and y axes an you can get a 'convincing fit' (good enough, often to convince a jury).GLD223 said: