Probability studies would say that the layouts are entirely possible, which they are. But personally I feel that pulling two or three cards from the previous day's readings and having them land in similar spots for two to five days in a row is obviously possible, but so unlikely that to say it isn't significant would be foolish. Lately I've just been doing a daily draw of a single card, but when I'm doing a daily spread I actually have gotten to the point that I find it more unusual to not get a repeated card or three (from an eight card spread). But yes, a probability study would most likely say that it was entirely random and merely a coincidence, no matter how unlikely. The only way that it couldn't follow under "possible" in probability would be to remove certain cards from the deck and have them show up in the spread anyway, or something else that had zero chance of occurring.
There is a statistical significance thing that shows certain events have such a rare chance of happening that if it does happen, it is significant. However, the nature of the subject means that if such a test has ever had such a result it would be dismissed by the scientific community as a fluke (because it's a rare chance but it is possible and therefore was a coincidence), and would likely never be published so we would never hear about it.
I found a way to test my intuition once using a five point scale for accuracy. As in, I'd interpret a spread for one card for each day of the week and record my impressions, then I'd journal my week in a place where I couldn't see the predictions (to not influence my writing). Naturally at the beginning of the week they were pretty clear in my head, but by about halfway through they started fading. At the end of the week I'd tally my results, using 0 for dead wrong and 5 for eerily accurate. Then I'd divide my total by the possible total, and that would be my percent of accuracy for the week. I ranged from 62% to 65%, and since I believe in free will and not destiny I'd say that was a pretty decent percentage for a sporadic student of the tarot.
There isn't something in there though that helps me figure out when I know I'm right. I mean, sometimes things are soft, and those things are possible but not set in stone. Sometimes things are hard and clear, and I know that's pretty much inevitable. Inevitable is super rare, but I only recall being sure something was super likely and it didn't happen, but it was dependent upon me having $3,000 (an opportunity) and I failed to gather the money. I thought I'd at least find out what it was that I couldn't do, but I never figured it out.
Anyway, my point is that except for that, when I feel sure something is definite, it's in the works already and my scale doesn't take how it feels like it will turn out into account, it only shows accuracy. It's also extremely subjective (self-reporting scales are already subjective and considered frequently to be flawed methods of gathering data, add that to the fact that even though I have not reminded myself of my predictions I obviously saw them at some point and am potentially being influenced by them, well, my study isn't likely to get published either).
It might be interesting to have someone do a weekly prediction for someone else's week, with that person keeping a journal and never seeing the cards laid out for them though.
