All major airline domestic flights departing from Oregon airports in April 2018
BA 275 Comprehensive Project
Project Idea: Using one data set (all major airline domestic flights departing from Oregon airports in April 2018), you will use all the tools we learn in this class. The point of any statistics class (including this one!) is to better understand the world through data. Although we have many tools, all of them come back to one question: “How can we better understand our data?” In order to really understand these tools, we will repeatedly ask the question, “what patterns can we find in this Oregon flight data?”
Project Procedure: Each week, you will be asked to use a different tool or approach on this same data. Each week you will add a new section to an ongoing project report. By the end of the course, you will have used every tool we have to better understand this data.
Project Data: the data is available on Canvas. If you want to check it for yourself, it’s available for free download here: https://www.transtats.bts.gov/DataIndex.asp
You may find the “glossary” to be of help in decoding the meanings of some of what you read: https://www.transtats.bts.gov/glossary.asp
Week 3: In class this week we’ve learned about tests for the difference of two means, which allow us compare two averages.
*Open your data in Excel or Statgraphics and answer the following in complete sentences.
1) Calculate a test for the difference of the average flight departure delay between Southwest Airlines and Alaska Airlines by selecting 10 random flights from each of these two carriers. Display your statistical output BUT NOT the “stat advisor”. Explain this result in a complete sentence (e.g. “because the p-value is (large/small), I (fail to reject/reject) the null hypothesis that Alaska and Southwest have the same average departure delay. In everyday terms, this means ______________.”
2) Now repeat the above test, but treat the whole data file as if it were a random sample of flights. In Statgraphics this is actually much easier than (1) above… In some ways. Select “Compare” from the top menu. Then select Two Samples -> Hypothesis Tests. You will be expected to enter in the sample statistics for each carrier (Southwest and Alaskan Airlines) but be sure to leave the null hypothesis as zero. We have no reason to assume the standard deviation of both groups are the same, so don’t make that assumption in the options.
3) Explain why the tests above are tests of independent (rather than matched) means.
4) How would we calculate a test of matched means between Alaska and Southwest? What would it mean for those flights to be “matched”?
5) Do you think both these tests are reasonable? Fair? Why or why not?