Math 256, Hallstone Term Project Fall 1999

advertisement
Math 256, Hallstone
Term Project
Spring 2008
Summary of Deadlines:
1. April 9 or 10 (or earlier)
2. April 17
3. May 1
4. May 15
5. May 29
6. June 9
Group names turned in using the last page of this handout, as well
as a meeting plan detailing how and when your group is able to
meet. Individual names must be turned in on the first date.
Plastic binder purchased, and everything placed inside. Main and
alternate hypotheses turned in, and your meeting plan turned in,
revised if necessary. All variables (including whether they are
categorical, discrete, or continuous) are identified. Each
hypothesis identified as involving one mean, two means, one
proportion or two proportions. Possible types of graphs that could
be used with your data are identified, and samples of what these
graphs might look like are included. Also, one copy of the Grade
Form turned in. I will return it and it will be turned in with each
deadline. Do not use sheet protectors at any time for this project!
Group meets with me to discuss your hypotheses. Corrections
made, if necessary, to type of variable and possible graphs.
First draft of -Part I and IIA turned in, as well as the original
hypotheses, and the Grade Form.
Data collected, second draft of Parts I and IIA turned in, and first
draft of IIB turned in, with all old work attached at the end. Also,
turn in the Grade Form.
Final draft of the whole project turned in, including all previous
work at the end as an appendix. Also a separate evaluation for
each person turned in (separately, if desired). Also, turn in the
Grade Form for your final grade.
This project is worth 11% of your course grade, and will be graded on a 100-point basis. It must be
worked on in groups of three or four students. The project in its final form will be due on the date
specified above. Note the preliminary dates as well. Each part/draft of the assignment that you get
back from me should be turned in again at the next deadlines, as part of the appendix.
The project is to be a set of hypothesis tests (tests of significance), and related statistics. That is, you
will come up with statements of the form: "I believe such-and-such to be true." Then you will attempt to
show by doing an experiment or by taking a sample whether or not they seem to be true. Hypothesis
testing takes many forms and you have the freedom to pick which type you would like. Some of these
we will cover extensively in class, but you are not limited to these and if the type you would like to do is
not covered, I can give you some help.
Your first step should be to come up with a main hypothesis, and a set of follow-up hypotheses. Here
are some of the main hypotheses that have been used before:
1. More than 50% of the people who smoke at GRCC do so outside of the designated areas.
2. The average age of books in the GRCC library is more than 10 years.
3. More people will wear seatbelts after a sign is put up than did before the sign.
4. The mean time to solve a puzzle is shorter for those listening to music.
1
I have a number of old projects in my office that you may look at and/or borrow. I also have a list of
some of the topics from previous years.
Some things are not good hypotheses for this type of project. You do not want to pick a hypothesis that
can be definitely verified. (For example, you could say "I believe that more than 75% of the people in
Math 256 this quarter are Business students." This could be easily verified or proved false by doing a
census of the students.) Instead, you want a hypothesis that requires a sample that you will use to make
inferences about the population.
By Deadline 1, I would like to know the members of your group. I would also like a plan showing when
you are able to meet as a group. (For example, you all have the 2-3pm hour free every day and plan to
meet at least every Thursday.) The biggest problem that people run into is choosing groups based on
friendships, rather than compatible schedules. Use the form on the last page of this handout.
By Deadline 2, I would like to get a short explanation of the main hypothesis you would like to test, and
what follow-up hypotheses you have. If you aren’t sure what to test, give me several alternatives that
you are considering. You may do observational or experimental types of projects. That is, you can ask
questions or observe some phenomenon, or you can actually do an experiment and record results.
Numbers 1-2 above are observational, whereas numbers 3-4 are experimental. Try to think about the
difficulties involved in collecting data for your particular hypothesis. (For example, if we changed the
easily verifiable hypothesis mentioned above to "I believe that more than 75% of all the people who take
Statistics are Business students", it would no longer have the problem of being easily verifiable.
However, it would now be impossible to collect such data, because you don't have access to that type of
information.) Also try to think of the interest level of your project. Will others want to read about your
result?
Also in deadline 2, I want you to identify your variables of interest and whether each variable is
categorical, discrete, or continuous. Decide whether your hypothesis involves one mean, two means,
one proportion or two proportions. See me if you are unclear about this; it is crucial to the project.
Finally, make up some data of the type you will collect, and make sample graphs of what you think it
will look like. (The purpose of this is to make sure you are very clear about what types of data you will
likely have.) Put all your materials in a 1-inch or 1.5 inch plastic binder, preferably with commercial or
hand made dividers.
I will return your ideas shortly after you turn them in and then you should then set up a meeting with me
to discuss how you will collect your data, and what follow-up hypotheses seem appropriate. (At or
before that meeting I also need any corrections made from deadline 2.) Before your discussion with me,
you should have a rough idea of how to proceed in terms of data collection. We will formalize this into
a plan. Note that data collection can be very tricky and very time consuming. (For example, to test the
"smoking at GRCC" claim above, you can't simply go to the student center (or the library, or group
members’ classrooms, etc) and ask 40 people whether they smoke outside the designated areas.) To test
hypotheses about campus, you will probably need to use a multistage sample, using randomness to pick
the time and locations. Very few projects will have an actual sampling frame for the population, so a
simple random sample is likely to be impossible. However, with proper coordination you should get a
sampling procedure that hopefully is still representative of the campus. This meeting with me must take
place before or during the Deadline 3 time frame. Make your appointment early to assure a time that is
mutually convenient. I will not be able to meet with every group on the last day!
2
When your project is turned in, it should be done on a word processor. It should have the following
format: (Note: Having this format does not mean that you should go through and answer each of these
questions. These are to give you ideas.)
I. Introduction.
What are your hypotheses? How did you come up with them? Why do you think they are important?
Why do you think the data will support these hypotheses? What are some of the implications if your
hypotheses are true? Have you seen similar reports or projects? This section should be 1-2 pages and
could be started as soon as your hypothesis is cleared with me. Keep a log sheet with the time and
place of each group meeting.
II. Data Collection.
A. Proposed. How do you plan on collecting your data? How will you try to achieve randomness? Do
you plan to use a random number table? Will you use simple random sampling, stratified sampling, or
some other method? Will you use surveys, polling, published material or something else? What
problems do you foresee? Will you eliminate as much bias as you can? How do you hope to deal with
your problems? This part should be done early in the quarter. I would like to have a preliminary word
processed copy of Part I and Part IIA handed in by Deadline 4. This should include the original
hypothesis you turned in with my comments. I will return this to you so that you may make revisions
before the next draft is handed in. However, you will not get full credit for this deadline if your writing
is poor. You should have your draft proof read by all group members and, if possible, by someone
outside the group. I highly recommend the Writing Center. This preliminary copy must go in the
appendix of the final form, so be sure to save it. You will lose points for each day this preliminary copy
is late, and lose points if it is not included in the final draft.
B. Actual. What did you actually do? How well did it go? What problems did/didn't arise? Were you
able to deal with them? Did you have to change the format of your hypothesis and/or did you have to
come up with a completely new one? Part II should be 2-4 pages. Remember the data must be collected
early enough that you can turn in word-processed versions of Parts I, IIA, and IIB by Deadline 5. This
will be returned to you in time to include it in your appendix.
III. Analysis of Data.
This is where you actually do the analysis of data and the statistics of the hypothesis test. Be sure to
check with me early, if you are not sure which type of test you should be doing. Which graphical
representations (pie chart, bar chart, histogram, boxplot, etc.) best describe your data? Which tests did
you use? Why were these the appropriate tests? Were all the guidelines and assumptions for using the
test met? (For example, is it a test for which you must assume normality? Did you need to check
guidelines like np0 > 10 and nq0 > 10, if applicable? Note that many of you will not need to assume
normality, but many will need to check np0, nq0. (Why?) What were the results? What tests did you
use to follow up, strengthen, or modify your conclusions? Was a confidence interval appropriate (it
almost always is!), and if so what type did you use? Did you need to collect additional data? What
effect does significance level have in your conclusion? Are there other tests that would be appropriate
but are not feasible at this point in time? This part should be 2-6 pages, depending on the number of
graphs that were reasonable.
3
IV. Conclusion
Did your tests agree with your original hypothesis? (Your grade does not have any relation to the
answer to this question.) If not, why do you think it didn't? Would further tests perhaps bear out your
original claim? If you were to do the project over, are there things you would do differently? What
implications are there to your results? Is there some organization, business, or campus department that
should receive a copy of this? What, if anything, have you learned from this project? This part should
be 1-3 pages.
V. Appendix
If you have any raw data (such as tally sheets, random number tables, or questionnaires) it should be
included here. Also, you must include all previous drafts and related materials.
Grading on this paper will depend on a number of things. (See the attached form that I will fill out for
each group.) Like any paper, you will be graded on the originality and importance of your ideas, your
proper use of the English language (including grammar and spelling), and the neatness and
professionalism of your report. Note that nearly every paper I see loses a number of points because of
poor English. I would advise you to have someone proofread the project before it is turned in. In
addition, you will be graded on how well you do at following the above format. I expect to see that
enough work was put in by all the group members. Each member of the group will do a self-evaluation,
as well as an evaluation of the other members of the group. A copy of the questions that will be used is
attached. These scores count toward your grade, so be thoughtful. If you have disagreements on the
evaluation about who helped on what parts, that will impact individual scores. I may ask group
members to justify their self/group evaluation. Also, you may be asked to give a short explanation of
your project to the class or me on the day it is turned in. Finally, your statistics are expected to be
sound. If you make a statistical mistake, it will hurt your grade very badly. (It is somewhat of a no-win
situation: good statistics do not guarantee a perfect grade, but bad statistics guarantee a lower grade.)
You will also be graded on the way that you make your paper readable to the average person. One of
the important skills of a statistician is the ability to converse with an outside world in a language that it
understands. Obviously, there are technical aspects to this paper and you can't make everything clear to
a lay person, but do your best. (The paper will likely be read by a second, non-statistical person as well
as by me. You may wish to have your proofreading done by someone with a good English background,
but not a statistical background. Once again, I highly recommend the Writing Center.) You will
have a chance to review my comments on your project, but you will not get it back; thus, if you want to
keep a copy, you should make one for each member.
Remember that procrastination could be detrimental to your grade, as there will be no extensions. You
must turn in your project by class time on or before Deadline 6. Also page 5 of this handout is the
self/group evaluation form and one must be turned in by each member at that time as well. However,
you may turn the evaluations in separately if you wish. Make sure you turn in the whole project,
including the final revisions of work previously turned in. Also put all the draft work in the appendix.
Don't forget that this project is worth 11% of your grade. You will lose points automatically if you miss
any of the deadlines.
4
Self/Group Evaluation Form
Project Title: ____________________________
List all the names, putting your name in blank A.
Group Members:
A. (This is you!)
_______________________
B. _______________________
C. _______________________
D. _______________________
1.
What, if anything, did you learn about statistics from this project? (Use the back if necessary)
2.
What did you learn about group work from this project?
3.
What are the strengths of your project?
4.
What are the weaknesses of your project?
Using the letters A-D from above to represent each group member, rate each group member (YOU are in space A!!) for each
statement (including the last one) with a number between 0 and 5 (5 being high).
Contributed greatly to the success of the project.
A
___
B
___
C
___
D
___
Contributed to the picking of the original idea.
___
___
___
___
Did much of the writing for the project.
___
___
___
___
Did much of the typing for the project.
___
___
___
___
Produced many of the graphs.
___
___
___
___
Kept us on task.
___
___
___
___
Created and checked the statistical tests.
___
___
___
___
Attended and contributed at group meetings.
___
___
___
___
Other: _____________________________
___
___
___
___
___
___
___
___
___
___
___
___
Overall grade for each group member. (Use 0-5)
(Remember, you are in blank A!)
(Not 0-5 on this one) Actual number of group meetings attended.
5
6
Grade Form (100 points possible): (This needs to be included in your binder with each deadline, except Deadline 1.)
Project Title: _______________________________________
Group Members:
A. ________________________
B. ________________________
C. ________________________
D. ________________________
1.
Deadline 2: [5 points]
1) Hypotheses, clear and complete meeting plan, and Grade Form turned in (with the binder).
2) Originality and appropriateness of your hypotheses
3) All variables are identified, and the types are correct.
4) Correct identification of 1 or 2 means, 1 or 2 proportions.
5) Possible graphs are identified, and good quality sample graphs are given.
6) No plastic sheet protectors. (Remember this for all deadlines!)
2.
Deadline 3: [5 points]
1) At least some of the group members met with me by the deadline, bringing this form with them.
2) Corrections have been made to type of variable and sample graphs, if necessary.
3.
Deadline 4: [10 points]
1) Quality of the first draft of Parts I, IIA
2) Grade Form and original of hypotheses turned in
3) Previous work is in an appendix at the end
4.
Deadline 5: [10 points]
1) Quality of the second draft of Parts I, IIA (This should lead off your paper at each deadline.)
2) Quality of the first draft on Part IIB
3) Grade Form and originals of previous parts turned in, but previous work is in the appendix.
5.
Deadline 6: [5 points]
1) Quality of the second and third drafts of earlier material.
2) Quality of the new material
3) Final paper looks and reads like a single, cohesive, and professional project.
4) Grade form and originals of previous parts turned in.
6.
Appropriateness and clarity of the data collection process.
3.
Relevance and completeness of graphical and numerical summaries of the data. [10 points]
4.
Final statistical analysis of the data. [15 points]
1) Appropriate tests and confidence intervals used
2) Tests done correctly
2) Proper guidelines and rules of thumb checked
3) Appropriate conclusions made, based on your data.
5.
Style, appearance, grammar, and spelling.
7.
Readability and interest level for someone not in the statistics field.
8.
Self/group evaluation (done by each individual). [10 points]
1) Thoughtfully done
2) Agreement as to who did what.
[10 points]
[10 points]
A. ________________________
C. ________________________
[10 points]
B. ________________________
D. ________________________
7
8
If you have formed a group, this form is to be turned in to me by April 10th at the latest. Turn in one
per group. I would prefer to get it back as soon as you have a group that has compatible times to meet,
so I can assist other groups to form. If you are not in a group, please turn in the first two parts
(name and all your free times) individually by April 9th or sooner. Note that this is one day earlier
than the group deadline, so I can match you up.
Group Names:
1. __________________________________
2. __________________________________
3. __________________________________
4. __________________________________
We are a compatible group because we are all free on the following days and times:
______________________________ (fill in days and times; for example, Mon-Thurs 10:00-10:30).
______________________________
______________________________
We understand that there is work outside of class for this project and we will be able to meet when we
need to. Here is our plan of when and where we shall meet to work on this project: (For example, We
will meet for at least 10 minutes every MWF at 10:00. As due dates come closer we will meet from 1011. We are planning to meet from 9-12 on In-service Day. )
Plan: (You will lose points if this next part is not completed! It is not sufficient to fill in the blanks
above saying when you are free.)
9
Download