There is no accepted way to determine cause vs. effect of an arc site. I recall Vyto doing a review of the various techniques that supposedly answered the question of cause vs. effect. His conclusion was that you might as well flip a coin; nobody had a process that was correct much more than 50% of the time.

An article was published in the latest issue of the Fire and Arson Investigator titled "A Review of the Long-Standing Science Behind Arc Melting Identification." Some references in that article may be of use.

The biggest thing you could do is consider the arc site in the context of the scene. What were the available fuels? Where were they in relation to the arc site?