Project Assignment for Statistics 371 Professor Wardrop Summer Semester 2011 You may perform the project described below and submit a written report for it Reports must be submitted on paper not electronically You may work alone or as a member of a team of three or fewer persons If you work as a team only one report should be submitted by the team remember to include everyone s name on the report and all members of the team will receive the same grade Neither my expectations for nor grading of reports is influenced by the size of the team Please submit legible reports For many of you this will mean words typed and nonwords e g tables figures formulas and computations drawn by hand Effective use of color and illustrations can greatly enhance a report The maximum score for the project is 2 5 points extra credit I have 42 students thus I could receive as many as 42 reports Sometimes reports disappear Some of those times it is due to my negligence In any event you must keep a copy of your report in case the original is lost Project Due before 12 01 pm on Thursday July 28 Perform 100 dichotomous trials Your trials might be Bernoulli Trials and they might not be Bernoulli Trials Part of your fun will be to investigate this issue see below We will discuss possible choices for trials in lecture Your report should describe the activity that generated the trials and you should explain briefly why this activity interests you I am not looking for originality but I want to know why the activity is interesting to you Here are two examples 1 I have always loved to play basketball so I want to study my shooting Free throws 3 point shots etc 2 My friends and I play quarters to see who must drink Diet Snapple My trials will be attempts to bounce a quarter into a juice glass Note that if a success is extremely difficult or extremely easy to obtain 100 trials are not likely to reveal much of interest You will obtain more interesting data if you select conditions that make the probability of success on a trial be a number between 0 25 and 0 75 a Before you collect any data make a point guess of the value of p Point means single number For example if I were shooting free throws before I collect data I would say My best guess is that p 0 53 b Use the data from your 100 trials to investigate the validity of Assumptions 2 and 3 of Bernoulli trials using the techniques presented in Course Notes Chapter 6 Note I put this in bold face type because historically a large proportion of students have neglected to do this when told with regular face type Your report must contain the time order results of your 100 trials so that I can check your analysis Let s make this clear To check assumption two you should create the first half second half table discuss what it reveals and do Fisher s Test To check assumption three you should create the memory table discuss what it reveals and do Fisher s Test c On the assumption you have Bernoulli trials use the standard normal curve to obtain the approximate two sided 95 percent confidence interval for p Is your point guess from part a in your CI What does the CI tell you about your guess 1 d On the assumption you have Bernoulli trials find the point prediction and 95 percent prediction interval for the number of successes in 100 future trials e Perform the 100 additional trials and compare your results to your predictions from part d Note I do not want you to check the assumptions of BT s with this second set of data Grading of Reports Here is a guide to how the projects will be graded Basically I look for errors If an error is particularly egregious I will count it as two or perhaps three errors Egregious errors are rare and will occur only if you neglect to do some big part of the project or you totally mess up some big part of the project Occasionally I will erase some errors if your project is wonderful in some way But don t aim for wonderful it is more productive to avoid errors The table below shows the relationship between the extra credit points you will receive and your total number of errors Number of Errors Extra Credit Points 0 or 1 2 5 2 or 3 2 0 4 or 5 1 5 6 or 7 1 0 8 or 9 0 5 10 or more 0 0 Very important read this Based on a problem that arose recently I want to make the following very clear 1 If you share data with another individual team you will receive a score of 0 on your project 2 If you use data from a report submitted previously by a student in any of my classes you will receive a score of 0 on your project 3 If you fabricate your data you will receive a score of 0 on your project Apart from issues of integrity and honesty here is my point I assign these projects so that you can get the experience of doing a study and so that you have the possibility of seeing that Statistics might be useful in your life By the way it is my judgment that will determine whether any of the forbidden actions listed has occurred In particular I don t need to prove it in any legal or statistical sense but I will of course be willing to discuss the matter with you if you think I am mistaken Model projects are available on my webpage Model projects 4 and 5 are examples of your project assignment Be careful in using these models Do not simply rewrite one of them plugging in your numbers and treatments trials This is because the specific questions I ask you to answer in the description above do not exactly match what I asked previous students Need some ideas On my webpage click onto Research Papers Papers on Statistical Education Small Student See examples 25 30 on pages 8 and 9 for ideas for project topics Or talk to me about it 2
View Full Document
Unlocking...