Monitoring panel performances with g p p the Mixed Assessor Model ...
Monitoring panel performances with g p p the Mixed Assessor Model ...
Monitoring panel performances with g p p the Mixed Assessor Model ...
Create successful ePaper yourself
Turn your PDF publications into a flip-book with our unique Google optimized e-Paper software.
The <strong>Mixed</strong> <strong>Assessor</strong> <strong>Model</strong> (M. A.M.)<br />
The M.A.M. C.A.P Table<br />
Application to <strong>the</strong> SensoBase<br />
The scaling effect<br />
Panellist<br />
Panel<br />
Disagreement<br />
A B C A B C<br />
A A B B C C<br />
‐4 ‐2 5<br />
However <strong>the</strong> <strong>panel</strong>list seems<br />
to agree <strong>with</strong> <strong>the</strong> <strong>panel</strong>. He<br />
just uses a larger scale.<br />
How to get rid of <strong>the</strong> scaling<br />
effect <br />
We search <strong>the</strong> number by<br />
which we multiply <strong>the</strong> scores<br />
of <strong>panel</strong> to be as near as<br />
possible of <strong>the</strong> <strong>panel</strong>list<br />
scores.<br />
2 is <strong>the</strong> coefficient of scaling.<br />
‐2 ‐10 3<br />
Strong disagreement<br />
X 2<br />
Pure disagreement<br />
Panel<br />
Disagreement<br />
A B C AA BB C C<br />
‐4 ‐2 0 6<br />
3<br />
Université Sensometrics de Bourgogne, 2012 le 20 Janvier 2012<br />
<strong>Monitoring</strong> <strong>panel</strong> <strong>performances</strong> <strong>with</strong> <strong>the</strong> M.A.M. 5/15