I am usually very stubborn about whats true or not. But if something happens consistently, it can be measured consistently right, so if science doesn't find anything, that's a sign its not true. Its as simple as that. Like, they could test you and others like you saying they are talking to siri on their phone, and then put the phone down and ask you not to speak to her, but to see if you can sense the evil. Then you will be like yes yeah yes, and then they will be like, tricked ya! There is no siri on the phone, I lied. Then you will be like 'cursesssss' to the heavens lol. Half joking here. Or they could have 5 phones one does have Siri they prove it, then they have you leave the room so you don't see how they shuffle the phones, then you need to pick the one with siri on it. Boom tricked again! Cursesssss! Siri wasn't on any of them lul. So here is a simple study you can do right. A person takes a deck of playing cards, they put just one with a very lightly placed symbol of something holy, then another with a symbol of something evil. Then, you turn around they shuffled, then they place all the cards out, none have dents from the pen cuz thatd ruin the test, then you see which one is good and which is bad. If you do not get it its because you just imagined it, and its easy to test. That will prove it is psychological priming like I mentioned earlier. But to be safe statistically speaking you want to do it 30 or more times and record the results. You cannot change your mind about the symbols half way in if you get it wrong, they have to remain good or bad, and there should be a small quantity of them, like 1/52 is good and 1/52 is bad, so its alot more obvious and quick to reach results that are correct if you are correct, and to not redefine your results after as its cheating and a defense mechanism of coping.
And the friend should shuffle it so they do not know where the cards are either. And shuffle thoroughly, and with a new set of cards or damages or bends could tell you.
That is more sciencey, testing to find out and making sure it doesn't suffer problems that influence the outcome horribly. Like if you had to just shuffle 3 coins you could track it auditorily or with your eyes when they do so, even subconsciously, and there is a chance you get the odds higher as its 1/3 and if you say do 1/2 of the time it seems like its confirmed. Also body language of a friend could reveal answers and its easy to track if its 1/3 for example. Or a coin could be less shiny than its counter-parts or a different year. It has to be uniform.