Z-test? Paired t-test? T-test two sample assuming equal variances?


New Member
Hey guys!

I'm hoping this is the correct forum to post this in. I'm working on my senior thesis, and it's been a while since I've taken stats so I'm hoping you guys can help out.

I'm looking at firms following an event and finding the return on their stock price in a set time period. I'm also finding the return on the market over that same time period. My null hypothesis is that the mean return on a stock price following an event is equal to the mean return on the market over the same time period. I'm pretty sure this will be disproven.

Anyways, I'm having difficulty choosing which test I should use. Like I said, it's been a while since stats. I'm leaning towards the z-test since I have a sample size greater than 30 (70). If I do a t-test, I'm really not sure which test I should be using. I have to believe there will be some correlation between the market movements and stock price, which leads me to want to use a paired t-test.

Any insight would be greatly appreciated!

I'm also having trouble interpreting results. I'm running it in Excel, and again am rusty.

So, for one of my z-tests I received :
Mean1 : -.53(Stock Return) Mean2: -.038 (Market Return)
z = -2.68
P(Z<=z) one tail = 0.003
P(Z<=z) two tail = 0.007

I really can't remember how to express this in words. I know that it is significant, and I think it means I can say with 99.3% certainty that the return on Stock Price is less than that of the return on the Market, but I'm not sure. I'm also not sure how to interpret the one tail vs two tail.