I’m preparing a controlled user case experiment for my thesis research in which I’ll give the users a definition to use.
The definition is not huge but it has some complex structures inside which I don’t want the users to manipulate during the experiment (I know they can and will be tempted).
I started clustering the components and saw the clustering increased the computation time by A LOT.
- is it a normal consequence of clustering?
- I don’t need to instance the clusters, indeed they are all unique, so is the another way of encapsulating the components without decreasing the performance?
Performance and reaction times are crucial to me because they will likely affect the user perception in the experiment.
Any help would be very much appreciated!!