The iodine differences are fairly consistent, but there are some random differences mixed in with the bunch (but that could just be operator error). I can try re-standardizing the thiosulfate solution I have to see if the molarity has changed much.
It seems to me that if that is what the problem is, my results would have a consistent pattern (i.e. the amount of iodine titrated would constantly be decreasing, do to a constant change in the molarity of the titrant). What I mean is, if the initial molarity of the titrant was 0.1 (I dilute to get 0.0005M), it would either decrease in molarity, or it would increase. But it wouldn't increase, then decrease, than increase again. Right now, the amount of iodine I am titrating now is the same (nearly) as it was last week, and the week before.
Is my logic right?
As for adding the starch solution later, I would do that (that's what I do in the thiosulfate standardization), but the amounts of iodine I am titrating are so low (I am using a 10 mL buret, and I only add about 0.3 mL of titrant), that I don't think I'd visibly be able to see the color change without the starch solution.
As for the starch solution, does it matter what starch I use? I've read all sorts of literature/ textbooks, and nothing is very specific.