add subesection for how to chose smd parameter

30_Thesis/sections/60_evaluation.tex

@@ -94,6 +94,15 @@ These user profiles can be used to generate rather homogenous groups but also to
 \section{Results}
 \label{fig:Evaluation:Results}
 
+\subsection{Choosing Happiness and Unhappiness Parameter}
+
+Results for happiness increase will vary depending on the chosen $smd$ value. This is to be expected because if it is set to low, the number of people that are happy with the decision of an individual will increase. An increase in amount of happiness means, that there is less people that can additionally be made happy. In the data though even with an $smd = 5\%$, this expected effect cannot yet be seen. But it is noticeable that the amount of people happy with the individual decision is decreasing with an increased $smd$. If $smd$ is set too high the group compromise will cause less people to be satisfied. This is because with the individual decision there will be always one person that is perfectly happy. Clearly this has the effect that choosing an $smd$ that is too high the decision maker will be satisfied with her own decision but no one else will. Additionally no one of the group will be happy with the group decision. Therefore it is expected, that with a high $smd$ the change in happiness will reach a value of negative one. The data shows these effects already with an $smd$ of $25$ where across the board all scoring functions result in a happiness decrease for the group (when looking at heterogeneous groups) but they don't reach levels of minus one in the tested range ($smd \in [5,35]$).
+
+Surprisingly, all tested values for $smd \in [5,35]$ resulted in a decrease of unhappiness. But as expected the total number of unhappy decreased with an increase of the $smd$, as was observable already with happiness.
+
+\missingfigure{Figure showing happiness and unhappiness with individual decision in relation to smd}
+
+
 \todo[inline]{explaining evaluations}
 
 \missingfigure{Result figure}