2
$\begingroup$

I was playing around with making 3D scatter plots, and needed some data to play with. So I took the results from Oct. Self-Evaluation, but I scored it a different way. I took the net score from the eval ("Rescore") and compared it to the votes of linked question ("Q Vote"), and the votes of the highest voted answer ("A vote").

That gave me the following data table:

\begin{array} {|r|r|r|r|} \hline Question &A ~ Vote &C ~Vote & Rescore \\ \hline Q1 &1 &1 &-1 \\ \hline Q2 &9 &9 &4 \\ \hline Q3 &3 &3 &7 \\ \hline Q4 &1 &3 &-2 \\ \hline Q5 &2 &2 &0 \\ \hline Q6 &8 &3 &8 \\ \hline Q7 &1 &2 &-1 \\ \hline Q8 &3 &3 &-2 \\ \hline Q9 &2 &2 &2 \\ \hline Q10 &2 &1 &-6 \\ \hline \end{array}

Which graphically looks like this (negative Rescore in red): graph of self eval

If I have time, I will go back and look at the other self evaluations and see if I can get any patterns to emerge.

But even with these data points, I'm re-evaluating my voting, though I know it has been discussed in depth. Are we voting in a way that reflects the quality we are or are not bringing to the internet? Should we be voting in such a way that it would be reflected in the Rescore?

I didn't have time to participate in the review, so I'm not entirely sure how the questions were selected (I'm assuming randomly of positive voted questions in the last quarter), but I would be curious how this would play out on the site as a whole (which none of us have time to do).

One last graph, the best liner plane fit that I could get in R:

liner regression of 3d data from self review

Just some data for thought.

$\endgroup$
2
  • $\begingroup$ In the first graph, Q10 is mislabelled as Q9 (two of those). Nice to see the results in graphs, but would probably be more useful if we combined many evaluations (and maybe colour coded different types of questions). Generally, I also suspect that people voting on evaluations do a more critical evaluation than "ordinary" voters. $\endgroup$
    – fileunderwater
    Commented Nov 4, 2013 at 10:23
  • $\begingroup$ @fileunderwater Sorry about the labeling, I'm still figuring out how do to effective 3d graphing. I plan on compiling the data from the other ones during pointless meetings this week. Of note, there are only two points significantly off the regression: [Q3][1] and [Q10][2] [1: biology.stackexchange.com/questions/9902/…: biology.stackexchange.com/questions/10176/… $\endgroup$
    – Atl LED
    Commented Nov 4, 2013 at 14:24

0

Reset to default

You must log in to answer this question.

Browse other questions tagged .