Wednesday, November 18, 2009
Monday, November 16, 2009
Tuesday, November 10, 2009
Wednesday, November 04, 2009
Monday, November 02, 2009
Tuesday, October 27, 2009
More statcrunch details
http://rchscrunch.wikispaces.com/
and see what is still available. Be aware! The page numbers may have shifted a bit.
After you pick your data set, edit the page and put your pick on the page!
That's right, it is a wiki: that means you can edit it. Click "Edit", type in your data set to claim it. Then click "Save".
I will be available this Thursday and next Tuesday in a computer lab for help.
Good luck!
Monday, October 26, 2009
last week and today
Tonight:
Describe how you would take a survey to assess the motivation that students have to get a Renaissance shirt.
How would you collect the data?
What biases would you worry about?
What is the population?
Wednesday, October 21, 2009
report
Goal: To show that you understand how to make and describe various
categorical and quantitative graphs and can use those graphs to
discover relationships in a large data set.
Output: A StatCrunch report that is completed and shared/emailed with
me and our RCHS statcrunch group.
Deadlines:
October 26th, Monday: Pick your data set and report it to me.
Note: your data must come from page _____ and can be chosen from the
top/bottom half. Go to Explore—Data to find your page. If that page
does not contain sufficient data, add 35 to your page number.
November 6th, Friday, 7am: Report must be completed, an email sent to
me and shared with RCHS group.
A few details:
*Categorical/quantitative graphs are both required.
*Scatterplots are not required, but are encouraged if they are
appropriate for your data. Likewise with regression equations.
*Your goal is find relationships in the data and describe them. The
most common way to do this is to graph one variable with respect to
another (movie revenue according to rating, pulse rate by gender,
etc…) However, graphs of just a single variable may be important as
well.
*This will count as a test grade.
*I will have a tutoring day in a computer lab for tech fearful!
Monday, October 19, 2009
Thursday, October 15, 2009
End of homecoming week
Due on Monday
Photograph on Monday
Friday: Fix test
Monday: Final questions
Tuesday: Test on:
Ch. 10: transformations
Linear Regression
Normal
Wednesday, October 14, 2009
Wednesday, October 07, 2009
Review answers
b) 233.5: For each year that goes by we predict about 233.5 more aircraft flying.
c) 89.9% of the variation in aircraft flying is explained by regression on year.
d) Plug in 2 (not 1992!): predicted aircraft = 2939.9+233.5*2 = 3406.9 aircraft.
e) residual at 2 = 40, so the actual was 3406.9+40=3446.9 or 3447
f) sqrt(.899) = .948 = strong, positive, linear relationship between year and aircraft.
g) On average, my predicted number of aircraft misses by about 33.43
h) aircraft-hat = 2939.93 + 233.5(year)
i) we predict in 1990 that the number of aircraft flying is about 2939.9
Study hard! I hope this helps!!!
Tuesday, October 06, 2009
Thursday, September 24, 2009
Regression practice
#28 all + (e) interpret slope
#33 all + (f) interpret r + (g) interpret r^2
Monday, September 21, 2009
Thursday, September 17, 2009
Tuesday, September 15, 2009
some review ideas, Unit 1
Q1 - 1.5IQR
Q3 + 1.5IQR
sd is always positive
(or zero if all the data is identical)
Categorical graphs:
bar and pie
Quantitative graphs:
stem and leaf (shows data)
boxplot (can hide shape)
ogive/cumulative freq
dotplot
histogram
Independent data makes bar graphs with similar levels
The mean is lower than the median for skewed left data
Measures of spread: IQR, sd and range
Measures of center: mean and median
Which of the above are resistant? med and IQR
Rule of thumb: 68-95-99.7
z-score formula: (x - mu)/sigma
Monday, September 14, 2009
Wednesday, September 09, 2009
normal cdf
Answers:
a) Only 0.07% of tires last longer than 40,000 miles. Not very likely!
b) 21.19% (z score is -0.8)
c) 67.31% (z scores are -0.8 and 1.2)
Tuesday night's HW
Page 113 #36
If you ever come looking for homework and I forgot post it, feel free to email me! I check email on my phone too much, so you'll get a fast reply!
Tuesday, September 01, 2009
Monday, August 31, 2009
Review and test
Monday: Unit 1 review, page 105 #7, 16, 21, 29 a-e, 31 and 32
Tuesday: practice, practice. Tutoring
Wednesday: Jeopardy?!?
Thursday: Test (no tutoring)
Friday: StatCrunch in the library computer lab
3-day weekend!!!
Thursday, August 27, 2009
Wednesday, August 26, 2009
Tuesday, August 25, 2009
Sunday, August 23, 2009
StatCrunch this week
- Go to My Groups
- Click on RCHS group
- Click on the First Day of school data
- Make a graph of one or more variables on the survey
- Describe
- Print graph and description
- Find some data. Lots of it.
- You have to find data that contains both categorical and quantitative variables on the individuals.
- Go to My Data
- On the left, you'll see some options, choose the one that works.
- Make sure to Share with our group:
- If you are viewing the data in the StatCrunch, click Edit and then you can share.
- There is a cool new option that will grab data from a website
- Go to http://www.statcrunch.com/
bookmarklet - Go to that page and read about adding a bookmark called "StatcrunchThis".
- Then go to the website you want to grab.
- Click the bookmark
- Save and data and share it.
Thursday, August 20, 2009
Wednesday, August 19, 2009
data, data
Ch. 4 #9, 12, 17, 24
Don't forget: if technology fails you, just make a dotplot!
Or you could try making a graph on Statcrunch!
#24 is a problem with statistics done WRONG. You task is to figure out what is wrong.
Tuesday, August 18, 2009
Monday, August 17, 2009
Thursday, August 13, 2009
independent?
In class we did Ch. 3 #17 and 18
For homework: Ch. 3 #15
Monday, August 10, 2009
Saturday, August 08, 2009
Stat Crunch instructions
Thursday, May 07, 2009
Competition Information
Friday, May 01, 2009
resources
Last weekend!
Thursday, April 30, 2009
Friday review session
Wednesday, April 29, 2009
Tuesday, April 28, 2009
Monday, April 27, 2009
Web site for Name that Test!
However, I think that after you 10 or so of these, you'll get used to the wording of this site and it will be good practice.
Note: this website includes prediction intervals, which you do not need to know.
Monday, April 20, 2009
Friday, April 17, 2009
Thursday, April 16, 2009
Wednesday, April 15, 2009
Tuesday, April 14, 2009
Friday, April 10, 2009
probability
Thursday, April 09, 2009
Wednesday, April 08, 2009
Review so far
Friday, April 03, 2009
Thursday, April 02, 2009
a little of this, a little of that...
Tuesday, March 31, 2009
Tuesday, March 17, 2009
CLT video
This should help you think about the Central Limit Theorem
Monday, March 16, 2009
Thursday, March 12, 2009
practice, practice
The X^2 for the evens are: 3.677, 290.131, 479.508 (listed in no particular order!). But that lets you check your work, at least a little bit!
Wednesday, March 11, 2009
Tuesday, March 10, 2009
Tuesday, March 03, 2009
Thursday, February 26, 2009
Wednesday, February 25, 2009
Monday, February 23, 2009
Thursday, February 19, 2009
Tuesday, February 17, 2009
Chapter 21 recap
The bigger the sample size, the more power we have.
We also lower our chances of making either error.
More Sample Size!
We need to be careful with REALLY big sample sizes.
The standard deviation gets really small and you'll always get a small p-value.
This can cloud practical significance: use a confidence interval instead!
Ho is false!
If the Ho is really false, we have lots of power and very little Type
II error (dumb criminal).
If the true value differs only slightly from the Ho, we have little
power and may make a Type II error. (smart criminal)
This is called effect size.
Changing alpha
Making alpha smaller (0.01) makes it harder to reject.
We do this if want less Type I, but are willing to tolerate more Type
II and less power.
i.e., death penalty
Alpha and medicine
If the established treatment is highly effective, you demand a very
small p-value before rejecting (low alpha).
If the standard treatment is not very effective and you have reason to
believe it can be easily improved, you would be more willing to accept
a new treatment even though the p-value is larger (higher alpha).