Type II error is the probability that the study doesn't find a difference between treatment and placebo (i.e. doesn't find a significant effect of treatment) but one actually does exist.  This is usually due to small sample size.