t stands for t test, and z stands for z test.
t is a kind of distribution similar to z distribution, while z is normal distribution(gaussion distribution), for a sample that is big enough, there is little difference between t value and z value.
if the sample is too small ,the z test will not be so reliable as t test
in a word, if t or z is big, it means that your hypothesis is wrong... for example, if your hypothesis is that the distribution of a random variable is normal distribution with mean u and standard variation delta , then you test a random variable , if the t is big (bigger than 2), then seems that your hypothesis is wrong--- the mean may be bigger. with high probability