A previous question inquired about two experimental groups, both made up of college freshman. A paired t-test was deemed appropriate for the two related samples. A follow-up question was asked:

Although the samples areactually related, an investigator ignores this fact in thestatisticalanalysis and uses a t test for two independent samples. How will this mistake affect the probability of a type II error?

*Edit:

The question is more of a discussion question, there is no actual data that was collected. It is asking me to explain what doing the experiment as two independent samples, instead of a paired t-test would affect the probability of a type II error.