Investigating approximate statistical properties of scaled problem instances of Simpson's Paradox using big data cluster computing techniques

This paper uses big data cluster computing techniques to investigate the approximate statistical properties of scaled problem instances of Simpson's Paradox. A visualization is presented for Simpson's Paradox. Although the... [ view full abstract ]

This paper uses big data cluster computing techniques to investigate the approximate statistical properties of scaled problem instances of Simpson's Paradox. A visualization is presented for Simpson's Paradox. Although the problem is intractable, big data cluster computing techniques are used to determine the instances of ball combinations checked and the instances of Simpson's Paradox found and ratios determined and plotted. This can be used to help approximate and understand the asymptotic behavior of Simpson's Paradox.