help
help > Graph Theory - significant nodes
Jul 31, 2019 01:07 PM | Davide Fedeli
Graph Theory - significant nodes
Dear Conn experts,
We are performing a graph theory analysis on our dataset and we have investigated the effect of a continuous variable on the clustering coefficient metric.
Our results show a significant effect on a network node (p = 0.02 with FDR correction) but not on the whole network (p = 0.52).
How can we interprete this result? Can we report that the effect does not impact on the whole network but significantly affects only a network node?
A second question concerns if it would be theoretically possible to test a three nodes network. If yes, how can we chose an adequate threshold, since the data fully overlap both with random and lattice graphs when selecting cost as threshold (see attached file)?
Thank you for your kind help and support
Davide
We are performing a graph theory analysis on our dataset and we have investigated the effect of a continuous variable on the clustering coefficient metric.
Our results show a significant effect on a network node (p = 0.02 with FDR correction) but not on the whole network (p = 0.52).
How can we interprete this result? Can we report that the effect does not impact on the whole network but significantly affects only a network node?
A second question concerns if it would be theoretically possible to test a three nodes network. If yes, how can we chose an adequate threshold, since the data fully overlap both with random and lattice graphs when selecting cost as threshold (see attached file)?
Thank you for your kind help and support
Davide