Announcements
- December 10, 2002 (4:24 AM)
- I've posted some more information about the exam on
my
final exam information web page.
The new information is in bright green. My
current guess is that I won't add any more comments about the exam
(unless I spot something that's wrong on my web page). I'll now turn my
attention to trying to create an exam that matches the description I've
provided, and isn't too hard --- I plan to make some of the questions
very easy in order to have an exam that you shouldn't feel so rushed to
complete.
- December 8, 2002 (6:44 PM)
- I've posted some information about the exam and my availability on
my
final exam information web page.
- December 5, 2002 (5:29 PM)
- Since our last class was cancelled, and it doesn't seem reasonable
to try to make it up, I'll alter my plan for the final exam to exclude
what I had intended to cover tonight. I'll make the HW exercises that
were previously assigned, but based on the material that I needed to
finish up tonight, into bonus exercises, having any earned points
contribute to the numerator, but not having the point values contribute
to the denominator. I'll try to post some information about the final
exam soon, and then I'll add more details to the study suggestions I
give once I firm up the specific questions/problems.
The exam will be based on material related to Chapters 3 through 9 of
H&W. (It's too bad that
our last class meeting was cancelled, since I wanted to tie up a few
loose ends from previous meetings. However, the material in Chapters 10
and 11 of H&W is not mainstream nonparametric statistics, and so we did
in fact cover all of the crucial parts of H&W (and some not so crucial
parts as well).)
- December 4 (my birthday (don't ask)), 2002 (4:21 AM)
- I've posted HW solutions through Exercise 22
on the
solutions web page.
I also updated the
reading guide.
The forecast
for Thursday calls for snow. If it does snow, please check official GMU
sources (as opposed to counting on me to inform you) to determine if classes have been cancelled --- unless the
university cancels the evening classes, I plan to present the last
lecture of the semester on Thursday. If we do get snowed out, I'll have
to make adjustments to what I had planned for the final exam. Because I
think there is a good possibility of a cancellation, and I don't think
we can make up the class unless the university designates an official
make-up day, I'll hold off until after Thursday night with making up
a small study guide for the final. Then, shortly after Thursday, I'll
post some hints about what to prepare for on this web site.
- November 21, 2002 (4:44 AM)
- I've posted the last HW exercises that I'll assign you this
semester on the
homework web page. (These last
exercises should not take long to do.)
- November 20, 2002 (2:04 AM)
- I've posted some comments about Ch. 10 on my new
comments about Ch. 10 of H&W web
page. I'll add more comments later.
I also updated the
reading guide.
- November 14, 2002 (5:55 AM)
- I added some more comments about Sec. 9.6 on my
comments about Ch. 9 of H&W web
page. The newly added comments are in green
so that you can easily
spot them. Also, I posted some more HW exercises on the
homework web page. These newly posted
exercises, and the ones which I will post next week, will be due at the time of the
final exam. You can note that I made the point values very small for
the parts that you may have to do using Minitab on the mainframe,
so that you won't be drastically penalized if you should choose not to
bother with them.
- November 10, 2002 (10:39 AM)
- I
updated the
reading guide, and posted some
more
comments about Ch. 9 of H&W.
(Not really related to the course, but since I know that at least two
people from STAT 657 attended my 11/1 seminar on the work I did for the
GMU basketball coach, I'll add some links to some related web pages that
I created.
This one gives some performance measures
for individual players, and
this one contains a description of how my
(currently incomplete) system for measuring a player's overall value
to the team works.)
- November 7, 2002 (7:02 PM)
- I forgot to add to my announcements of early this morning, that
last night I also corrected my information on my
StatXact web page about doing
the one-way layout version of Page's test (using the Linear-by-Linear
Association test).
- November 7, 2002 (5:22 AM)
- I
added to the
homework web page, and
the homework solutions web page. I
also
posted some comments about Sec. 8.5 on my
comments about Ch. 8 of H&W web
page.
- November 4, 2002 (1:15 PM)
- I
updated the
reading guide, and posted some
comments about Ch. 9 of H&W.
- October 26, 2002 (2:55 AM)
- On Friday, Nov. 1, at 10:45 AM, I'll be giving a seminar about some of
the analysis that I did for the GMU men's basketball coach. It'll be a
somewhat
light talk, but perhaps more interesting than most of the talks in the
Friday morning seminar series. Here is the
abstract for my seminar. Someone recently inquired about homework
scores, and so I decided to tally up the points at this time
(through what I returned on Thursday, and so not including the
problems that were due on Thursday (since I haven't graded those yet)).
I summed everyone's points so far, and used the formula presented
here to produce a value on a
scale of 0 to 60 for each student. (The scale being based on 60 points
is due to the fact that HW counts 60% and the final exam 40%.) To apply
the formula, I used m = 85 (even though it would have been
possible for someone to have earned 90 points so far because of 5 extra
credit points (that count for the numerator, but not the denominator)).
Due to my scheme of making an allowance for missing/skipping some parts
of the HW, one only needs 90% or more of the 85 points (or at least 76.5
points) to achieve the maximum possible HW score so far. Six of the 12
students who are taking the course for a grade are currently at 60 (the
maximum possible score), and the other 6 scores (rounded to the
nearest integer in some cases) are 59, 55, 52, 49, 44, and 42. For
those of you who aren't doing as well as the others, I suggest that you
try to work on the HW exercises well enough in advance of when they are
due so that you can ask questions about whatever uncertainties that you may
have about the proper procedures. It is my hope that I can give higher
than usual letter grades for this course, like I did this past summer.
- October 24, 2002 (5:32 AM)
- I posted a bunch of new HW problems on the
homework web page.
Please note that for a test based on a multiple comparison procedure,
there is just one "overall" p-value (and not k choose 2
different p-values). Ask me about this if you don't understand.
- October 24, 2002 (1:27 AM)
- Even though nobody asked for any additional information, I posted
some comments about the test corresponding to part (b) of Exercise 15 on
my
StatXact hints web page.
- October 23, 2002 (5:30 AM)
- I posted some more solutions on the
homework solutions web page.
I also posted some comments pertaining to Exercise 16, Exercise 17, and
Exercise 18 on the
homework web page. In particular,
for Exercise 16, I added a reminder that you're supposed to report two
approximate
p-values for each part; one based on not using a continuity correction,
and one based on using a continuity correction.
I'll add that I'm starting to like H&W a bit more now, in part because
it presents some useful tests that one doesn't see covered in too many
places.
- October 22, 2002 (4:24 AM)
- I completed my
comments about Ch. 7 of H&W.
I've also updated the
reading guide, adding some
information at the very bottom of the page (as it now exisits)
giving some loose indication of how extensively I plan to cover some of
the sections in Ch. 7. (Note: We may not finish going over Ch. 7 in
class this week since it is a rather long chapter. If this is the case,
I'll cut short the time I had planned to spend on Ch. 10, and otherwise
allocate the class time as I've indicated on the
syllabus.) In order to help make
sure everyone understands the notation and general approach, I plan to
go through a small example and calculate the value of the Doksum test
statistic. I encourage you to take the small set of values below and
compute the value of the Doksum test statistic before coming to class.
|
treatment 1 |
treatment 2 |
treatment 3 |
block 1 |
45 |
140 |
235 |
block 2 |
50 |
106 |
202 |
block 3 |
39 |
76 |
60 |
block 4 |
70 |
220 |
373 |
block 5 |
10 |
24 |
44 |
- October 21, 2002 (4:44 PM)
- I added to my
comments about Ch. 7 of H&W,
covering through Sec. 7.9. I plan to continue adding to my Ch. 7
comments some more tonight. I've also updated the
reading guide, adding some
information at the very bottom of the page (as it now exisits)
giving some loose indication of how extensively I plan to cover some of
the sections in Ch. 7.
- October 21, 2002 (6:25 AM)
- I added to my
comments about Ch. 7 of H&W,
covering through Sec. 7.7. I plan to continue adding to my Ch. 7
comments as I work throughout the day.
- October 20, 2002 (4:50 AM)
- Saturday afternoon at about 3 PM, my e-mail mailbox got full
due to way too much spam, and I didn't get it taken care of until about
8 PM. So if you tried to send e-mail to me between about 3 and 8 PM on
Saturday, you may very well need to resend it. (I think at this point I
have no STAT 657 e-mail that I haven't responded to, so if you're
waiting to get a reply from me, you had better e-mail me again.) By the
way, be sure to always put STAT 657 in the subject line of any
e-mail message that you send to me so that I can better identify it as
nonspam. (The spam problem is horrible. An old address like mine has
made its way onto many lists by now. Filters prevent some spam from
getting through, but spammers seem to always come up with schemes to get
things past filters, and so I tend to get about 4 spam messages per
hour. So if I don't get to my mailbox for over 24 hours, and especially
if some student sends me an inappropriately large file, my mailbox can
get full and no new e-mail can be received. Since on some days spam
outnumbers legit e-mail at about a 10:1 ratio, particularly on the
weekends, you need to remember to clearly identify yourself as a STAT
657 student in the subject line of any e-mail message that you send.)
Also, an observant student caught a mistake I made on my solution to
Exercise 10, in the part dealing with the normal scores test. My silly
mistake didn't cause the approximate p-value to be different from what
it should have been when rounded to 2 significant digits, but the exact
p-value was off by 1 in the 2nd significant digit. I don't think anyone
had points taken off because of my mistake, but if you had 0.012 for
part (c) and seemed to not get full credit, turn your paper back in so
that I can make a proper adjustment.
I also updated the
reading guide.
- October 17, 2002 (6:25 PM)
- I added one more HW Exercise (to be due in 2 weeks) to the
homework web page.
- October 17, 2002 (5:20 AM)
- I added one more HW solution to the
homework solutions web page, and I
posted one new HW Exercise on the
homework web page. I'll report that
for the last bunch of HW exercises that I graded (25 points worth), the
class median is 20 and the class mean is about 19.54.
- October 16, 2002 (6:36 AM)
- I added some more HW solutions to the
homework solutions web page over
the past two nights.
- October 15, 2002 (12:55 AM)
- I added some comments about Exercise 12 to the
homework web page. Hopefully, the
comments make it more clear what I am looking for.
- October 14, 2002 (12:31 PM)
- I added some comments about Exercise 13 to the
homework web page. I'll try to add
some other information about the homework tonight or tomorrow.
- October 14, 2002 (9:19 AM)
- I added to my
comments about Ch. 7 of H&W,
covering through Sec. 7.4 (which is how far I expect to get
in class this week).
I also updated the
reading guide. and added to the
StatXact web page. I plan to post
some comments related to the homework later this morning, so check back
later and perhaps you'll find my comments to be helpful if you don't
know for sure what is expected.
- October 13, 2002 (5:31 AM)
- I posted some comments about Sec. 7.1 of H&W on my
comments about Ch. 7 of H&W
web page. Also, I'll let you know that one of the people in the class
let me know that one can indeed copy and paste on the Case Data sheet.
Just highlight the desired data using the mouse, and use the Edit menu
to copy --- then click on where you want to put the copied part and use
the Edit menu to paste. (I was trying to use the left mouse
button to copy and paste, but that doesn't work.)
- October 10, 2002 (4:52 AM)
- I added to my
comments about Ch. 6 of H&W.
I also updated the
reading guide. and added to my
StatXact hints web page
(which I also added to yesterday, but forgot to put in the
announcements).
When grading homework this past weekend, I realized
that I had forgotten to record the scores on the papers that I returned
to you last time. I have the values, because I had computed things like
the mean score, but I didn't put any names with the values. So could
you please bring your graded paper with you to class and let me get the
score off of it, or just e-mail me your score if you have your paper
handy by the computer at home or work?
- October 9, 2002 (5:22 AM)
- I added to my
comments about Ch. 6 of H&W,
and plan to add more comments later (like late Wed. night or early
Thurs. morning). I also posted some more exercises on the
homework web page.
- October 6, 2002 (12:46 AM)
- I posted homework solutions to Exercises 4 through 7 on the
solutions web page. The scores on
the second set of exercises were pretty good for the most part, with 8
of 12 students getting at least 27.5 (out of 30 --- so better than 90%).
The class median is 28.25, and the class mean is about 25.17. Since you
always have at least 2 weeks for the HW exercises (except at the end of
the semester when you will only have one week for whatever exercises are
assigned during the last week of class before the exam), I encourage you
to try to do each exercise by one week after it is assigned so that you
can ask about whatever you discover that you don't fully understand.
- October 3, 2002 (5:46 PM)
- I decided to make Exercise 2 (that I graded and plan to return to you
in class tonight) an extra credit problem (which basically means
that the points you earn from it count towards your total points earned,
but I don't count the 5 points it is worth in the "denominator sum").
I decided to do this before I graded the papers because I realized that
I didn't talk to you very much about the confidence interval the problem
dealt with. But I was pleased to see that many of you got it correct!
All in all, the scores for the first 3 exercises (out of 10 points, since
the points you earn are the extra credit exercise (Exercise 2) are pure
"bonus" points) were very good:
- 5 of 12 students got a 15,
- all but 1 student got at least a 10,
- the class median is 13,
- the class mean is 11.625.
I edited the
homework web page and
this web page to reflect the
extra credit status of Exercise 2 and explain how the extra credit
points will be handled. I also added one more exercise to the
homework web page.
- October 3, 2002 (6:15 AM)
- I posted some homework solutions on the
solutions web page, and plan to
return some graded work to you tonight (but I haven't gotten everything
graded yet).
Since I had hoped to give you feedback on the
second set of problems that you turned in before you had to submit the
problems due tonight, I'll let you turn in the problems that were to be due
tonight next Thursday (with the extended deadline being the following
Monday). (This way, I can return your graded homework to you on
10/10 and you can study the solutions to them and turn in your work by
the following Monday if you aren't confident in what you have done.)
I also updated the
homework web page to show the
problem I'll distribute in class tonight, and I've created a web page
giving
StatXact hints that I'll add
to as the semester continues.
- October 2, 2002 (6:00 AM)
- I posted some
comments about Ch. 6 of H&W,
and also updated the
reading guide.
- September 28, 2002 (2:52 AM)
- I updated the
homework web page, created a web
page giving a
Minitab macro for the
jackknife test of Sec. 5.2 of H&W, and fixed some minor typos and
errors in grammer that "Eagle Eye" reported to me (thanks, Kathy).
- September 26, 2002 (5:57 AM)
- I added some comments about Sec. 5.2 to my
comments about Ch. 5 of H&W
web page. (I found Ch. 5 to be the most interesting chapter in H&W so
far. I hope that you enjoyed it as well.)
- September 25, 2002 (5:55 AM)
- I posted some
comments about Ch. 5 of H&W,
but haven't added any comments about Sec. 5.2 yet. I also edited the
reading guide
(but only made a couple of small changes).
- September 23, 2002 (5:25 AM)
- I updated the reading guide.
I'll post some comments about Ch. 5 of the text over the next 48 hours.
- September 17, 2002 (10:24 AM)
- I added comments about Sec. 4.2 to my webpage which gives
comments about Ch. 4 of H&W,
which completes the comments that I had planned to post for this chapter (although
there is a possibility that I may decide to add some more comments
later).
- September 17, 2002 (7:43 AM)
- I added to my
comments about Ch. 4 of H&W,
but I still have to add some comments about Sec. 4.2. I also updated
the
reading guide.
Yesterday I
corrected two minor typos that a student spotted, and I also came up
with a new color scheme for my web pages about the chapters of the text,
so that if you print them you can get easy-to-read black text on a white
background.
- September 16, 2002 (11:07 AM)
- I started typing in some
comments about Ch. 4 of H&W
in the wee hours of the
morning and I will continue adding to what I have so far throughout the
day. I encourage you to use StatXact to obtain p-values for some
of the examples from H&W, e.g., Example 4.1 on pp. 110-111. (My
comments about Ch. 4 of H&W
provide you with some instructions pertaining to this example.)
- September 12, 2002 (5:26 AM)
- I added 4 more homework exercises (due Sep. 26 & Oct. 3) to the
homework web page.
- September 11, 2002 (3:33 AM)
- I added more
comments about Ch. 3 of H&W,
for Sections 3.7 through 3.11. I'll discuss some of the posted comments
in class this week, as well as present some other material related to
Ch. 3. I also updated the
reading guide.
- September 9, 2002 (5:03 AM)
- I didn't get around to posting my comments about the rest of Ch. 3
yet, but I'll try to do it soon. (I'm not going to have near as many
comments about the last part of Ch. 3 as I did about the first part of
Ch. 3.) As you read the rest of Ch. 3, don't be too concerned about the
details of the rarely-used tests presented in Sections 3.9 and 3.10 ---
I certainly don't plan to emphasize those tests. I encourage you to get
started on the HW, and to prepare some questions to ask about the first
3 chapters, if you find that there are parts that you don't understand.
Finally,
I'll report to you that the StatXact order was placed on Friday,
but since the person in charge of shipping wasn't working on Friday
(StatXact is produced by a small company), the order won't be
mailed out until today, and it may not arrive in time for me to
distribute the software during class this week. (Since 10 people
ordered the heavy two volume manual set, it would cost too much to have the
shipment sent using overnight delivery. (The department secretary has
already scolded me for not adding any shipping fee to the prices I gave
you.)
- September 3, 2002 (5:48 AM)
- I posted some homework exercises (not due until Sep. 19 & Sep. 26) on
the homework web page. I may add
more exercises to be due on Sep. 26 if the StatXact shipment
arrives. Remember to bring your check with you on
Thursday if you want to order any StatXact stuff.
- September 3, 2002 (2:50 AM)
- I added more
comments about Ch. 3 of H&W,
either finishing, or coming close to finishing, what I wanted to post
prior to this week's lecture. Also, for user convenience,
I added links to my comments about
various chapters of H&W at the bottom of
this web site's home page.
- September 2, 2002 (6:02 AM)
- I made some
comments about Ch. 3 of H&W.
(The comments pertain to the first part of the reading from Ch. 3 that
you're supposed to do prior to this week's lecture. I'll add some more
comments later, but most of the more invloved comments pertain to the
first part of this week's reading.)
- August 31, 2002 (8:29 PM)
- I made a couple of very minor changes to
my comments about Ch. 2 of H&W,
but nothing of importance. (Thanks again go to "Eagle Eye" Harrison,
for spotting a minor mistake (I omitted the word is in one place).
Please point out such little things to me. Even though I may not use
H&W again, I'd still like to have a nice set of web page comments about
each chapter of the book.) Below are a few more comments for today.
- Since I'll guess that many of you didn't read the material (the
text, and my web site comments) for the
first lecture, please do so (in addtion to reading the material for the
second lecture), and ask questions about it during office hours or
in class this coming
Thursday.
- Here is the
abstract
for my Friday the 13th seminar. Since I'm only supposed to talk for
at most one hour that Friday morning, I don't think I'll be able to add
too much to what I presented to use last Thursday night, and so there
wouldn't be a lot of point in you making a special trip to campus for my
9/13 talk. (Later in the semester, I'll give a talk about data analysis I did to
help the GMU basketball coach. Since you won't have already heard that
presentation, you might want to try to attend that Friday morning
seminar.)
- Both the Provost and an Associate Dean have urged the faculty to
urge students to check their GMU e-mail accounts regularly, since those
e-mail addresses will be used to distribute important information from
time to time. It's possible to have your GMU e-mail forwarded to
another account of your choosing, and so maybe some of you will want to
do that. Since I plan to use my web site for communication with you, I
will only send e-mail if something out of the ordinary occurs and I want
to increase the possibility that you will get a message from me (like,
for example, if Robinson Hall burns down and class has to be cancelled).
In such cases, in addition to posting information on my web site,
I will send an e-mail message to whatever addreess(es)
you wrote down on the index card you were supposed to have filled out
and given to me this past Thursday.
- August 30, 2002 (1:47 AM)
- I edited the
reading guide to show what you are
supposed to read before next week's class. On that web page you'll also
find links to my comments about the material covered by this week's
lecture, and I encourage you to read those comments if you haven't done
so already. You should note that my web site comments are intended to
complement the lecture, and not duplicate it. So there will be topics
covered in class that aren't in my web site notes, and there will be
comments on the web site that aren't made in class.
(Also, be sure to read the
syllabus and related web pages like
this one giving information about
HW if you haven't done so already.) I think that it might be a good
idea to lug the big H&W book to the class each time --- since I don't
plan to give out lecture notes, when I explain some of the details in
the text, you might want to have it to look at, and perhaps jot down a
note or two in the margins.
(Also, if you do plan to bring your book to class, you might want to put
the check for your StatXact purchase in the book, since you'll
want to be sure to give me your check next time so that you don't miss
out on the sweet deal for the software.)
- August 27, 2002 (3:57 AM)
- I edited my
web page giving information about
homework, adding comments about turning in HW late, and about the
presentation of HW solutions.
- August 27, 2002 (2:22 AM)
- I added
some comments about the Chapter 2 of
the text. During the first lecture, most of the time that I spend
on Ch. 2 will be related to Sec. 2.3, which deals with confidence
intervals.
- August 26, 2002 (4:28 AM)
- I added
some comments about the Preface of
the text, and
some comments about the Chapter 1 of
the text, and added links to these two web pages in the
reading guide.
- August 25, 2002 (3:45 PM)
- I added a comment to the Additional Comments section
towards the bottom of the
syllabus (see the 3rd bullet, about
putting STAT 657 in the subject line of e-mail messages).
- August 21, 2002 (1:29 AM)
- I added
a bit of information about the
homework component of your grade. I also corrected a few typos and
made some very minor changes here and there. (When I make very minor
changes, like fixing typos or changing fonts or colors, I won't give the
details here. But I will announce any important
changes.) Thanks to
Kathy Harrison
for finding my typos. (I'll find some of Kathy's mistakes later.)
- August 20, 2002 (5:42 AM)
- I have an extra copy of the text (new, never been used) that I will
sell to someone for an even $100.00 (cash --- I don't like to deal with
checks). Amazon.com and the Wiley web site both list the book for
$105.00, and I believe that was the price at the GMU bookstore. (Btw,
there are plently of the text books on the shelf at the bookstore. If
you don't want to save $5.00 by buying my extra copy, you might want to
come over to GMU before classes start next week and buy the book, since
the bookstore lines can get very long during the first week of classes.)
- August 20, 2002 (4:40 AM)
- Tonight I created the
syllabus (but I may add some
details about the homework and grading later).
- August 2, 2002
- Please read the information on the
welcome page and check back
later for the syllabus and other information.