Hi I am somewhat confused about something...I have read that ideally one wants to select a sample that is homogeneous. But doesn't this create problems of restricted range? Any info would be great!

not sure of the context but typically if you are trying to get a good idea of what is going on in a population you want a sample that is representative of the population. A homogeneous sample will give you the smallest variance and therefore the smallest confidence interval but that doesn't mean it is an accurate representation of your population which is you real goal. For example lets say you wanted to determine how many children each person in your county has given birth to but you want you sample to be homogeneous so you only sample men. Well each man has given birth to 0 children and you will have a very small confidence interval but you will also not have any better idea of how many children a person in your population has given birth to on average.

Okay, well let me give you an example from a study I am conducting. I am conducting an experiment looking at whether writing about a stressful life event will help ameliorate the negative impact of the event. I recruit individuals who have experienced this event and randomly assign them to either the experimental group or a control group (where they write about trivial topics). Now, it's my understanding that you would want your samples to be homogeneous because if not it introduces within group variability, but if your sample were homogeneous then wouldn't that result in a restricted range on your outcome measures?