Great question! I would change the over-reliance on instrumentation used to "prove" hauntings when there is no scientific proof that any of those instruments can do anything to detect a ghost. I'd also change the language we use and assumptions we make. For example, we assume ghosts are the spirits of dead people, but we don't know that. We don't know what causes the anomalies we notice – not really. I don't care if you see and record an apparition materializing right in front of your eyes that looks like your Great Aunt Fannie. Just because that is what you see and capture on film does not mean it is the spirit of Great Aunt Fannie. It may be, but it could be other things, as well.

These assumptions make me nuts, yet we use them all the time. And when we report our experiences, investigators rely on terminology that seems to confirm these assumptions. I seldom see findings presented without interpretation, either. Using the above example, in a report an investigator might say, "I saw the ghost of Great Aunt Fannie." That statement is riddled with interpretation. In fact, what you saw was a figure that appeared with the features and form of Aunt Fannie. We need to stop reporting interpretation as fact.

We assume EVPs are voices of the dead and present them as such. That's an interpretation. In fact, EVPs are merely audio anomalies we can't really explain. We do ourselves a disservice by reporting these as something we don't know they actually are. If we assume we know what they are without actually trying to figure it out, then we are exactly what critics call us – pseudoscientific.

That being said, the other thing that really bugs is the current trend moving away from personal experience. We are our own best tools. Our observations, our experiences and our feelings do matter on investigations – but we need to report them as just that – observations and not scientific fact. To do anything else obfuscates data and keeps us from making actual progress.