Week 5 (2/24/05 - 3/3/05)
Karl Pearson (1857-1936)
-- originator of many statistical methods, including chi-square and r
HW
- Collect HW3
- Return and discuss HW2
- Feedback is most effective when given early, so . . .
- Good job, and (but?) I want to see continued development in levels
of presentation and interpretation
- In the future, statistical reports, not mere scratchings, and certainly not raw computer
output
- According to the course syllabus (p. 2, #7) "Results should be reported neatly and
should conform to APA Publication Guidelines for text and numbers
when possible. Use judgement when reporting results."
- When asked to show work, please include statements of H0
and H1 (so I know you know what you are testing)
- Discussions should be more than "reject the null hypothesis"
or "means differ significantly." P values are NOT
enough. You discuss direction and magnitude of differences, spread (if n
moderate) outliers, etc.
- Calculations by hand, some by computer only, some by both, depending
on what was covered in class.
- Hand calculations may be checked by computer (e.g., Did anyone use the online
ANOVA calculator pointed out in Ex. 12.5B to check their calculations on
Ex. 12.6?
LECTURE CHAP 14 (CORRELATION)
- The main data set used to illustrate concepts in this weeks chapter may be
downloaded: bicycle.sav
- Since the pdf Makers seems to have been booby trapped by Microsoft, I'll
bring hard copies of presentation slides with me
LAB 4
EXERCISES - version
of 2/23/05
- 14.2 (Anscombe's quartet) -- no calculations
- 14.4 (na-bp) -- by hand (OK to check w/ computer)
- 14.8 (mat_mort.sav) -- by computer