The issues with EVPs are multiple, but if you consider that replies are not always applicable to the question, you have to wonder a few things:
1. Is it necessary to speak out loud? Ghosts hardly have an auditory canal.
2. What if the ghost is replying to something one of the many people in the room is thinking and not the question spoken out loud?
Narrowing the possibility that our scattered minds are confusing ghosts, let's try this experiment. Write down 5 questions. Start the session like this:
Leader with stopwatch calls out "Begin"
Everyone for 30 seconds thinks the first question. They read it in their head, concentrate on it, only think about that one question over and over again.
At 30 seconds, the leader says "next" and the concentrate 30 seconds on the next question.
They do this until all questions are done. Then, they stop the recording.
Now, correlate if answers are applicable when everyone concentrated on the same thought. We can make some assumptions based on this, though we can never eliminate the possibility of psychokinesis from one of the participants. We can learn if collective thought is more effective.
The next part of this experiment would be to see if one person is an activator. You would include everyone having the same 5 questions, except one person who will concentrate unknowingly on 5 separate questions. You will continue doing this pattern without telling people that one person will be having separate questions until each participant has been the odd man out. Now, see if any of those odd man out people were answered. If you have reason to believe this person might be a psychic sensitive attraction for the spirit world, attempt it one more time on that person without telling him/her and see if once again he/she gets answers.
There are lots of outside of the box experiments that can be attempted in the study of ghostly activity. I suggest researchers spend some time doing research in the field besides doing house calls. During house calls, do no try new experiments on your clients. Save these kinds of experiments for a private study of the team in a haunted location that does not include a client with complaints, such as a haunted abandoned jail or asylum, haunted B&B or historic site that allows studies.
If you do this experiment, please report back. I'd love to see how it goes for you.