PY603 - Advanced
232 Gordon Palmer Hall
|192 Gordon Palmer Hall
|408B Gordon Palmer Hall
|Tue 6pm on weeks when
homework is due
Announcements and Corrections
- 2007-12-06: Projects will be due on Wednesday, December 12. If
you need an extension let me know.
- 2007-12-06: Exam 2 will be posted on the website on Friday, and
should be turned in by Monday. It will cover material from Chapter 4 to
Chapter 9. The exam will be closed-book, so you should not look at your
notes, homeworks, prior exams, or anything related to this class
between the time you first open up the exam to the time you email it to
email@example.com. If you
are inadvertantly exposed to information related to the class during
this time you must contact me so that I can give you new questions that
are untainted by whatever information you have heard.
- 2007-12-06: Made a correction to the HW8 key. I had
misinterpreted the coefficients on my effect codes. Thanks Abbie for
- 2007-11-29: We'll be talking about some wise comments of Jacob
Cohen in Stat Lunch tomorrow. Come by 204 GP if you get a chance.
- 2007-11-15: Jamie, Mark Klinger, and Nancy Rhodes will be talking
about analyzing reaction time data tomorrow at Stat-Lunch. Come by 172
GP if you'd like to hear more!
- 2007-11-09: I've updated the regression notes to include new
sections on how to test the significance of specific simple slope
coefficients. The new content is located on pages 54-57.
- 2007-10-26: Avani will have limited availability between
Novermber 9 and November 21. She will not be holding the Stat Chat on
Tuesdays on November 13 and November 20, and may take a little more
time to respond to your questions.
- 2007-09-28: Brittany is taking over my job! She actually searched
and found the new link to Wooley's suppression article, which is posted
below. Guess it's time for me to retire to Jamaica and carve wind
chimes for a living.
- 2007-09-06: Jamie will be giving a presentation at Contemporary
Issues this week (Friday 9/7 at 1pm in 204 GP) entitled "Contemplations
on p-values." Please come if you have an interest!
- 2007-09-04: Avani's office hours have been officially changed to
Tuesday from 6-7pm. Feel free to contact her at other times as well if
that time doesn't work for you.
- 2007-09-04: I've added a new link up above to saved chat logs.
Whenever I have a Yahoo messenger chat with someone I will ask them if
it would be ok to post it on the website. When they do, I'll upload the
file into the chat directory so that anyone can take a look at it. We
will also save copies of Avani's office hour sessions to that directory.
- 2007-08-30: Below are some of the formulas relating to p-values
that you can find in Excel, along with explanations of how they work.
- TDIST(t statistic, df, tails): This formula gives you the
p-value for a t statistic and degrees of freedom. The t statistic can
be either positive or negative.
- TINV(probability, degrees of freedom): This formula will give
you the value from the t distribution where the two-tailed p-value is
equal to "probability." It always gives you a positive number. If we
label the result of the formula to be X, the probability is the
proportion of the t distribution that is less than -X plus the
proprotion of the distribution that is greater than X. If you want to
know the critical value when generating a confidence interval with
confidence level equal to Y, you would use the formula TINV(1-Y, df).
Therefore, if you wanted the critical value for a .95 confidence
interval that had 39 df, you would use the formula TINV(.05, 39).
- NORMSDIST(Z statistic): This formula gives you the proportion
of the Z distribution that is to the left of the Z statistic. Notice
that this is only the p-value when the Z statistic is negative and you
are interested in a 1-tailed test. If your Z statistic is positive,
then you have to either put -Z into the formula, or else subtract the
probability you get from 1 to get your p-value. For example, to get the
p-value for a Z of 3 you would need to use either the formula
NORMSDIST(-3) or the formula 1 - NORMSDIST(3). If you want the
two-tailed p-value, you would need to multiply this result by two.
Therefore, if you wanted the two-tailed p-value for a Z statistic of
2.5, you would use the formula 2*NORMSDIST(-2.5).
- NORMSINV(probability): This formula gives you value from the Z
distribution that is greater than the given probability. If the
probability is less than .5, then the resulting Z will be negative. If
the probability is greater than .5, then the resulting Z will be
positive. If you want to know the critical value when generating a
confidence interval with confidence level equal to Y, you would need to
use the formula NORMSINV(1-(1-Y)/2). Therefore, if you wanted the
critical value for a .95 confidence interval, you would use the formula
NORMSINV(1-(1-.95)/2), which would be equal to NORMSINV(.975).
- 2007-08-27: The notes on ANOVA and Excel that were posted prior
to today were actually older versions. I have updated the files on the
- 2007-08-27: Avani has changed the chat hours to be Tuesday at 5pm
because of student conflicts.
- 2007-08-23: Where does the yahoo pager ID come from? An old latin
sibili si ergo
Fortibuses in ero
O nobili demis trux
Si wat senem
- To WebCT login
- Related to data and data sets
- Related to multiple regression
- Related to testing model assumptions
- Related to mediation
- To other web resources