Launching PollQuant – A “Modeled” Poll of Polls
Orlando, FL, November 05, 2016 --(PR.com)-- Sid Som, owner and publisher of prominent data analytics and info sites like ChefQuant, Homequant, TownAnalyst and LocValu is pleased to announce the launch of PollQuant, a scientifically “modeled” poll of polls.
While a couple of poll of polls exists, they are average-based, without a scientific noise-reduction mechanism. Since PollQuant is modeled, it significantly reduces the incidences of noise and influences of outliers. When at least 15 major polling organizations are competing for visibility and recognition, polarizations would be expected -- for example, with one tracking poll registering +4 while another showing -10 for the same candidate during the same survey period -- thus negatively impacting the consistency and reliability of the average-based poll of poll.
Similarly, divergent sample sizes like 750 vs. 3200, unsound sampling bias leading to over/under sampling of likely voting groups, conflicting survey methods, e.g., direct telephonic contact vs. online polling, differing methodologies like moving averages vs. simple averages, etc. often make the component polls apples-to-oranges. Moreover, arithmetic mean-based calculations help perpetuate the influence of outlier outcomes, causing more volatility and instability in day-to-day movements.
On the other hand, when those competing polls’ attributes are modeled out in a single equation in a statistically significant manner, the component polls largely converge to the same results, laying the foundation for a more commensurable comparative platform and paving the way for more reliable and explainable outcomes. Furthermore, since the model realigns outcomes of outlier polls, no special judgment or treatment would be needed in handling or removing them before the final line-up is decided upon.
Though the traditional polls of polls are inherently backward-bending, PollQuant is not. Based on the concept and practice of predictive modeling, PollQuant produces results that could easily be used to forecast some short-term outcomes. In fact, it may not be an exaggeration to assume that when the poll pundits understand the underlying concept and usefulness of PollQuant, they will use it as a predictive tool to validate their own internal results.
In a nutshell, PollQuant serves multiple purposes. First and foremost, it’s the first statistically significant poll of polls. Secondly, it’s also an independent predictive poll, without the subjective nuances of an atomic poll, making it a perfect addition to a short line-up to strengthen its reliability. Finally, it could be used as a validation poll to authenticate one’s own internal polling results.
In a recent conversation, Sid Som, the inventor of PollQuant, stated, “With the introduction of our modeled poll of polls invention, we are adding quantitative science to an age-old market mechanism which was in desperate need of modernization. Now, pundits can comfortably rely on our poll of polls without having to worry if certain outlier polls skewing the overall metallurgy of the overly simplistic average-based outcomes. Similarly, our invention makes issues like sample illiquidity, bias, methodology, margin of error, etc. irrelevant. Our modeling process equalizes all of that, creating a truly commensurable comparative platform. I have no doubt that in the not-too-distant future our invention will force the poll of polls landscape to reinvent itself with a dose of science.”
http://pollquant.blogspot.com/
While a couple of poll of polls exists, they are average-based, without a scientific noise-reduction mechanism. Since PollQuant is modeled, it significantly reduces the incidences of noise and influences of outliers. When at least 15 major polling organizations are competing for visibility and recognition, polarizations would be expected -- for example, with one tracking poll registering +4 while another showing -10 for the same candidate during the same survey period -- thus negatively impacting the consistency and reliability of the average-based poll of poll.
Similarly, divergent sample sizes like 750 vs. 3200, unsound sampling bias leading to over/under sampling of likely voting groups, conflicting survey methods, e.g., direct telephonic contact vs. online polling, differing methodologies like moving averages vs. simple averages, etc. often make the component polls apples-to-oranges. Moreover, arithmetic mean-based calculations help perpetuate the influence of outlier outcomes, causing more volatility and instability in day-to-day movements.
On the other hand, when those competing polls’ attributes are modeled out in a single equation in a statistically significant manner, the component polls largely converge to the same results, laying the foundation for a more commensurable comparative platform and paving the way for more reliable and explainable outcomes. Furthermore, since the model realigns outcomes of outlier polls, no special judgment or treatment would be needed in handling or removing them before the final line-up is decided upon.
Though the traditional polls of polls are inherently backward-bending, PollQuant is not. Based on the concept and practice of predictive modeling, PollQuant produces results that could easily be used to forecast some short-term outcomes. In fact, it may not be an exaggeration to assume that when the poll pundits understand the underlying concept and usefulness of PollQuant, they will use it as a predictive tool to validate their own internal results.
In a nutshell, PollQuant serves multiple purposes. First and foremost, it’s the first statistically significant poll of polls. Secondly, it’s also an independent predictive poll, without the subjective nuances of an atomic poll, making it a perfect addition to a short line-up to strengthen its reliability. Finally, it could be used as a validation poll to authenticate one’s own internal polling results.
In a recent conversation, Sid Som, the inventor of PollQuant, stated, “With the introduction of our modeled poll of polls invention, we are adding quantitative science to an age-old market mechanism which was in desperate need of modernization. Now, pundits can comfortably rely on our poll of polls without having to worry if certain outlier polls skewing the overall metallurgy of the overly simplistic average-based outcomes. Similarly, our invention makes issues like sample illiquidity, bias, methodology, margin of error, etc. irrelevant. Our modeling process equalizes all of that, creating a truly commensurable comparative platform. I have no doubt that in the not-too-distant future our invention will force the poll of polls landscape to reinvent itself with a dose of science.”
http://pollquant.blogspot.com/
Contact
HomeQuant
Sid Som
718-314-4081
Sid Som
718-314-4081
Contact
Categories