Carson asks a question regarding something he saw on one of the TV Ghost hunter Shows.
To sum up your question in two words, "BAD SCIENCE." Loosening a flashlight end cap simply removes pressure needed to assure a good connection with the batteries. There is a reason why solid contact is essential for a light to operate. Batteries and terminals by nature may get a small amount of dirt or oxidation on their surface. This can easily disrupt the flow of current, or at the least cause a poor connection which can result in heating where the surfaces meet. Thus any slight disturbance will cause the light to react. Tight contacts assure this problem does not happen.
Two things can cause these disturbances. The first and most obvious is vibration. Even walking across a floor can result in slight vibrations passing to a flashlight laying nearby on the floor or a table. This can simply shake the battery enough to make or break contact blinking the light unless the battery is securely installed.
But the second disturbance is even more likely the cause. A poor connection has increased resistance to electric current, thus a higher voltage drop. Ohms Law related to power states that "Wattage is equal to voltage times current." Power represents heat and for this reason a poor connection will create heat. Heat results in expansion of the metal at the connection itself. This expansion results in movement, thereby causing erratic operation. Of course since the conditions are constantly changing due to heat the connection may at times cool if contact becomes more stable or heat as it becomes more resistive.
For this reason the light will alternately come on and go off. If you watch people use this method you will note that often the light also goes on and off when no questions are asked. Of course, since we are conditioned to only expect a response to a question these false positives are overlooked. But if a response coincides with a question we are expected to believe a spirit did it.
Making Bad Science Good
All is not lost however. There is a better way to attempt the same experiment. First consider the hypothesis itself. The theory is that a spirit can cause the light to turn on or off by applying pressure which controls the flow of current. If we follow that assumption, all that is really needed is a way to take the randomness out of the equation. For this we tighten up the poor connections and make a simple switch that, when pressure is applied, will turn on a light.
I should state before going on that the science presented here is proven; the spirit interaction is not. No one has even proven spirits exist, let alone that they can blink lights or answer questions. What is presented here is a means of testing the hypothesis not a proven method of contact. It is your responsibility to do the experiments and analyze the results objectively. The methods I cover only serve to minimize false positives caused by bad science and thereby improve credibility of any results you may obtain.
A basic switch assembly that anyone can put together is shown in figure 1. (Actual component values are not critical; I would recommend a single cell and a 1 1/2 volt flashlight bulb.) Here the poor connection has been replaced by a simple spring which acts as a switch. Just like the loose flashlight back it works by pressure closing the circuit causing a current flow and lighting the bulb. Only we have made sure there are no poor connections which cause false positives. The switch is closed by simply applying pressure to it, which is the same thing the spirit is supposed to do with the flashlight. An added advantage to this is that the switch could actually be placed some distance away from the observer thereby preventing accidental disturbances.
Making Good Science Better
One problem however is that we can never be sure that a spirit activated the switch. Maybe a strong air current pushed the spring down, completing the circuit. Or some other external force managed to disturb it. However it is a widely accepted theory that spirits or ghosts can pass through solid matter. So the obvious solution would be to place the above described switch in an enclosure. To improve the switch assembly simply place it inside a glass jar with a lid. Drill a small hole in the lid for the wires to come out and screw it on the jar. If in fact our spirit can pass through solid objects it should have no trouble reaching through the glass to activate the switch.
Making Better Still Better Yet
For this one we go high tech. A bulb does draw a bit of current and there is another theory that spirits don't actually close the switch. Rather they somehow allow current to flow through them. This is supposedly how spirits are supposed to interact with EMF meters or directly impress audio on recorders with no one actually hearing them. While this is also unproven, the method shown here could definitely put it to the test. The problem is that a weak spirit may not be able to allow sufficient current flow to light the bulb. We might compensate for that as well.
Figure 2 shows a simple solid state switch that will light an LED or bulb with minimal current through the switch. It uses a small current (microamperes) on the base of the transistor to control a larger current which lights the bulb. There are two metal screens positioned near, but not touching each other mounted on an insulating block of wood or plastic. This serves as the switch. It is so sensitive it may even be activated by a charge of static electricity in the air! Of course this could also present a problem with false activation, but it is a trade-off. You can minimize this possibility by placing the switch in the target area then leaving the immediate area. Any activation while in close proximity to anyone should be disregarded and considered false.
And to take it to the limits, consider placing the solid state switch inside the jar as stated above.!
These are just a few suggestions of simple experiments one may try to test this hypothesis. They make no assumptions spirits exist or can do any of these things. If you have any success to report from trying any of these ideas drop me an e-mail!