Lasted edited by Andrew Munsey, updated on June 14, 2016 at 9:13 pm.
• Congress:Top 100 Technologies -- RD (Main20)
• Congress:Bylaws (rd)
There was an error working with the wiki: Code
Standard Deviation Abberations in Voting for NEC's Top 100 Energy Technologies
We should establish an algorithms to see how far outside the "average tally" a given person's vote falls. If it is substantial, then the control panel could notify the person that their vote is substantially removed from the standard deviation, and that they might want to reconsider their stance. An exception would be if they voted "abstain" or "need more time."
Likewise, if a given person tends to vote far removed from the standard deviation in a lot of instances, that could be brought to the attention of an NEC membership committee for membership plausibility considerations. This would be ascertained in relation to other NEC members. The "top five" standard deviation mavericks would be noted for the NEC membership committee.
The program would then report the instances of standard deviation variance, with most severe listed first. Whether they consistently vote high or low, or a combination (all over the place), any of these might suggest incompetence or unsuitability.
Sometimes those who vote far outside the norm are among the most qualified members, and they have a significant reason for doing so. If the member thinks they have a reason for voting high/low that the other members are not considering, then they should present their case to the NEC so they can likewise adjust their votes. For this reason, a "comments" field should be available with each vote cast, so that such a reason may be supplied.
A similar algorithm could be run for the lay members participating in voting. If their votes fall sufficiently outside the standard deviation, then their "opinion" will be considered suspect. Either they are mindlessly entering information, or they don't know what they are doing and their voting privilege will be revoked, and/or their votes not counted.
In order to take such an action, though, there ought to be a minimum number of votes cast in order to have an adequate statistical sampling by which to determine sufficient variance.
Furthermore, there could be a feature evoked at the time a vote is cast. If there are at least a minimum number of votes by which to determine wide variance of the new vote cast, a notice could immediately be posted to the person, informing them that their vote is far removed from the average vote normally cast for that matter. They would be invited to provide a reason for their aberrant vote. If the vote is abnormally high, maybe they have insider information, e.g. they've seen the technology working. Their witness to that might then sway others to adjust their vote. If their vote is abnormally low, maybe they have tested the device or know someone who has, and the results were negative. Etc.
A variation of this algorithm would be a "vote spread factor". The less that is known about a technology, the more the vote might tend to sway high and low. Technologies with a wide vote spread should be noted as such, so that people know that the score is hazy. Technologies with a more narrow spread, likewise, should be noted as such with an appropriate label to the effect of "laser focus."
Congress:Technology Review Protocol - protocol for submitting and reviewing individual technologies.
- New Energy Congress main page