Discussion
- Data Distribution & Decision-Making The idea that the Gaussian
Distribution is so frequently found in everyday life that it is also called
Normal Distribution can also be misleading. If one assumes normality where it
does not exist, data-driven decisions will be in jeopardy. For instance, making
decisions that will affect a region with high unemployment rates under the
assumption that the average salary is the most common salary, and that there’s
an equal number of people who are well-off as there is
people struggling to make ends meet, would be catastrophic. For your initial
post, find one type of data that tends to be normally distributed, and one type
of data that does not. Make sure that you present relevant data/sources to
support your claims. Then, discuss how assumptions about the normality of
your “not normal” data can result in bad decision-making. In your follow-up
posts, help deepen understanding of your colleagues' cases, namely on the
claimed possible consequences of misperceptions
Probability distributions are pertinent aspects of describing
the behavior of populations. The reliability of the gaussian is undeniably
superior. However, the suitability of this distribution can be flawed in
various populations. Therefore, the gaussian distribution is juxtaposed against
the Pareto distribution in accurately explaining the behavior of distributions. The wide applicability of the gaussian
distribution has been instrumental in statistics and data science for a long
time. The notion that having a sample or population average and the standard
deviation of this average within the population can avail information about the
population distribution has been reliable, albeit now questionable (Hogg et
al., 2019). The internal structure of the members of a complex population is
rarely Gaussian distributed, and therefore, the applicability of the Pareto
distribution is more plausible for accurate population description.