Does anyone here have experience with statistics? I am getting annoyed by a problem in my healthcare statistics class regarding the difference between descriptive and inferential statistics and I'm looking for help understanding why I'm wrong (or possibly why the teacher is wrong).
My daughter is the statistics expert and unfortunately I am not. But ironically I had a discussion with somebody else--mostly related to political polling--and looked it up. This article seems to be very comprehensive, but in all honesty didn't help me a whole lot. (I didn't take a lot of time really studying it though.) Maybe it will make more sense to you?
Difference Between Descriptive and Inferential Statistics: Descriptive vs Inferential Statistics Compared
Unfortunately, the problem is that I have looked up the definitions of descriptive and inferential statistics over and over again, and I still am convinced that my teacher is wrong regarding a particular problem we had. At this point the teacher is annoyed and doesn't want to discuss it anymore, which is pretty frustrating because I ended up failing an assignment with only 6 questions because I got 2 wrong, but I still don't see how they can be wrong.
Basically, there's a sample of 500 people from New York City. 210 have O+ blood. Then you get a statement that 42% of the people of NYC have O+ blood, and are asked if that is descriptive or inferential statistics. I say it is inferential because it is taking a sample data set (the 500 people) and generalizing that to a larger population (NYC). My teacher is telling me that it is descriptive, because NYC is the data set and the 500 people are a sample of that data set.
What I've gotten from her is that inferential statistics only apply to a population outside of that which you draw a sample from, but that contradicts multiple definitions for inferential statistics I have seen. In the link you just gave, for example, it says "Inferential statistics is the branch of statistics, which derive conclusions about the concerned population from the data set obtained from a sample subjected to random, observational, and sampling variations." In the question, it seems to me that the "concerned population" would be the population of NYC.
I actually think that the questions may have been poorly written and that it may actually have been trying to say that 42% of the 500 people have O+ blood, rather than all of NYC.
Whatever the case, I'm not comfortable just leaving a failing grade when I think that I am correct. Unfortunately there doesn't seem to be a good way to prove (or disprove) my point. If I knew someone with a PhD in statistics I'd ask them, and maybe have them give me an explanation to tell the teacher if it turns out I'm correct.