The chart is nice, though we would be able to get a more accurate/valid idea had the total points scored in each category been included.
The "problem" I guess that I have, is with the fact that the ranking is being displayed in terms of the average score. IMHO, if rankings are to be done, then it would be more accurate to display the results in terms of total points scored, rather than by what each member "averaged". I realize what you're saying Stryfer. You can't really divide someone's scores by a certain number of voters if fewer people voted for her overall. It is because of this fact that I brought that issue up in the first place. Displaying/ranking them by their average scores just isn't showing accurate data, because the number of voters per member is different, which means they're not all being shown/organized on the same basis. Total points, on the other hand, are simple, straightforward, and are much harder to dispute.
Ayaka got 8 total votes, with a blank aka 0 in personality in three of them (5,5,5,5,3,0,0,0). So we divide the sum of those with the amount of positive votes (in this case 5).
...
And naturally if everyone participating in this vote gave her a zero for personality, the result would be the same.
I'm afraid I must disagree with you on that. In terms of the arithmetic, no, the results would NOT be the same. If you had included the 3 people who gave her a 0, Wada's average score would NOT be the same. By your own statement here, there were 3 people who submitted votes for Wada in any of the 5 categories. You did not include these 3 people in your calculations for her personality score because they gave her a score of 0 for it. In terms of the arithmetic, it is more accurate to divide her score by the total number of overall voters that she actually had, which is 8, and not only by the 5 you did. The 3 people who gave her a personality score of 0 still voted for her in other categories, correct? If so, then again for accuracy's sake they should all be included in the calculations for her scores in each of the five categories. You've given her total of 23 points, and divided them by 5 voters to give an average score of 4.6. Had you included the 3 people who gave her a 0, then it would have been a total score of 23 divided by 8 voters, giving her an average of roughly 2.8, which is a pretty big difference. For the displayed results to be accurate, they need to be organized on the same basis. Because of the fact that different numbers of people voted for different members (thus giving us some members with 20+ people submitting votes for them and others having 5 or less people submitting votes), it is simply not accurate to organize them by their "average" score.
If you take both Wada and Yuko's total scores, and divided each of them by 25 (i.e. the number of total voters that Yuko had), the comparisons between the results would have been more accurate because they would have both been divided by the same number of participant voters. Instead of "omitting" the "blank votes" that Wada
would have gotten, then Wada's "average" would have sunk like a stone, and she, like I said, would have placed much lower in the "rankings".
Let's put it this way, let's say I were to give a test/exam to a class, and there are 25 questions on it, each worth 1 mark. I told the kids "You only have to answer the questions you want to. If you don't know it, then you do NOT have to answer it, and you will NOT be penalized for it." Student A answers only 5 questions in total (leaving 20 blank), and gets 4 correct. Student B answers all 25 questions, and gets 20 correct. This gives me:
Student A - 4/5 = 80%
Student B - 20/25 = 80%
Now then, would it be valid to say that Student A and Student B should be ranked the same? Did they get the same score? Based on what questions they answered, both of their scores equaled 80%, even though one student obviously did more work and knows more of the material. As we all know, of course not. The accurate, and, more importantly, the fair thing to do would be to take Student A's score and mark it out of 25, because that's how much the test was out of. He/she did not know the material, which lead him/her to leave 20 blank questions. As such, he/she should not be lead to believe that he/she did well by omitting the questions that were left blank and only doing the ones he/she knew. Student A should not receive a score of 80%, he/she should receive a score of 16% (4 out of 25).
At any rate, I'm just sayin', is all. Like I said before, I'm not trying to start a fire here, but details like that just bug me sometimes.
Topic: Hello! Project Pentagon Poll [RESULTS P.2!! REACT] (Read 19318 times)