Math 151, Spring 2006 Day 3, Fri. Feb. 3,Hit reload to get most current versionAfter class

Day 3 (Fri. Feb. 3): Reading:  Reread D&V thru p. 18. (We'll come right back to pp. 18-24) Read Ch.4 thru p. 46  (Re-expressing p. 44 optional), + stemplot  handout. (Error in my book: Ch4p43 middle; "for men...a narrower broader peak...") Activstats 3-1, 4 all.
 Ahead, D&V Ch.3 pp. 18-22, 23-4 (Simpson's Paradox optional). Activstats 3-2.  Then D&V Ch5, AS Ch5 (D&V, and I, will do medians, quartiles, boxplots first, then mean/s.d.  AS does middles, then spreads, then boxplots.)

We'll start using SPSS Wednesday--have class in the computer lab that day.  Everything by "hand" till then!
Needed for HW: Stemplot, rounding when there are more than 2 decimal places?  Handout says truncate (round down), D&V text says round to nearest.  Tukey, the inventor, said truncate; throw away the trailing digits; I agree.  This is supposed to be fast--rounding to nearest slows it down.  I encourage truncating but you can do it either way and be right.  If you truncate, your stemplot may look a little different from the text answers. (A stemplot is hard for a computer to do, but some packages do. For them, rounding to nearest is easiest.  SPSS truncates, which is hard for a computer.)

Note:  Many people missed seeing that Ch 2 p.13,  4 Oscars, 5 Bears was part of Day 2 HW.  I've asked Fay to not count it late if it comes in Monday.
Day 3: Hand in Monday (all from D&V text)  
(Postponed problems are repeated here.)
Ch 3 p.28, TheBar/pie templates handout is helpful in creating the displays.
5 Death;
 12 Teen Tech II 
 13 Auditing [I could disagree w/answer to b]
Ch4 p 50
Creating:
12 bird species (10's as leaves, split 5 leaves per stem is good.  Big outliers)
18 Marijuana (stem &leaf)
 p.104, #1a, banana price Make a stem and leaf and a dotplot.  For the dotplot make a stack of each different number (e.g. a stack of 51's and a stack of 52's...) Which display do you like better?
Describing: 
 5 Heart attack stays
 9 Wineries: Make a) "under 60 acres". Book's answer to b is screwy, why?
 14 Pop. growth
More Ch4: #4 more shapes
A. Use your circle data and make a back-to-back stemplot of Time (first column) for your two hands.  Write a few sentences comparing the speed performance of your hands.
Read, be able to discuss in class
Ch3  11 TeenTech I
Ch4
Creating: 
 17 Acid rain Look at answer, note stems used 
Describing:
 7 Cereal sugar
 6 Emails (I think the answer book does a crummy job)
 19  Hosp. stays Do a only.  Read answer to c.  Most mothers & babies go home in 2 days now.  What W's are crucially omitted here? 
 
 


Optional 
 
 
 
 
 


Find someone you don't know (or who might not remember you) and (re) introduce yourself.
Don't forget to initial the sign-in clipboard.
Monday: here in classroom.  Wednesday:  Come to Computer Lab, Mac 101.  Bring text; disk or usb to save on (containing your circle data if possible.)
Class email list, Math151@wells.edu
Cluster  in 3's, 4's or 5's. Check for Homework questions? Remaining #s on board.
Each group fill in summary sheet of Circle colors/hand.  Pool separate results. (Hand Summaries in)
   Your questions
Pretests:  Mixed, mostly ok:  Order of op's-- Please Excuse My Dear Aunt Sally:  Parentheses rule; Exponents,  x, /, +, -.
   Take it to math clinic, anyone, ask for problems like the ones you missed.

Data:  Numbers  in context:  Who (are the cases & how many =n),What (&units),  When and Where,  How, Why?
   Context for height, hair color, shoe size, pulse rate:  Any problems with the way I did it?

Variable (possible values), individuals (cases)  recall
        Categorical (ordinal--has natural order or &&nominal--just names)
    or Quantitative (can add, average--measured on a ruler-type scale) Units?! ("calories"?)

OBJECTIVE (from syllabus): To learn many of the ways in which data can inform us about the world, focusing on

   "Circle experiment"
    -- how to present your individual data--so it answers your questions.
    -- design of the experiment: what things were done to eliminate "lurking" or "confounding" variables--that might mess up the data.
    -- inference:  Data represents what you DID.  Would you get similar results if you did it again?  Can we infer from our class results to other people?  The techniques and rules  for dong so.

Distribution of one variable:  Area represents proportion. ="Area principle"

  Categorical: (Relative) Frequency distribution-->Bar or pie graph 
                  (Bar chart ordered by size =&& "Pareto chart"--not in text)

         Pie &&only ok if showing all categories (part of whole) + no overlap of categories.
              .  Pie by hand?  Template handout
         Relative Frequency of colors in circle experiment?  Any patterns?  Variability?

   Quantitative: Histogram, Stem-and-leaf ("Stemplot"), Dotplot
      (I will only require you to read, not make histograms by hand. You'll Make stemplots and dotplots by hand)
  Histogram bar area represents the number (or proportion) of individuals (cases) in the interval ("bin") at the bottom of the bar.   "Living Histograms" handed around. 
   (Bar graph (categories) =>  Spaces between bars.  Histogram=> no spaces between.)
       Pretest:  Restate #5 as histogram of 100 "6-volt" batteries tested for actual voltage.
          20 between 0&1 volt, 10 between 1&2,  30 between 2 & 3, 30 between 3&4, 10 between 4&5.
              The proportion with voltage < 1 is 20%.  The proportion with voltage < 3 is 60%.
               a) What proportion have voltage beween 1 and 3?  b) What proportion have voltage > 3?

   Stem-and-Leafs are a powerful hand tool.  Handout
            Unordered first, then ordered if necessary.  By tens, then split?  (Ex.:Class data)
        Back to back, comparing two groups. (p.51, #14)

Choosing a display (by hand):
    A dot plot (p. 39) is most useful for n = 3 to about 15-20, or when the data only fall on a few values (just make your scale axis & stack the dots up).
    A stemplot is good for continuous data, smeared around; you can do 100 values in 3-5 minutes.

Read the rest, read the text, do the HW.  We'll go over this briefly on Monday
Describing:
  Pattern-- and deviations from it

   Shape (symmetric, or skewed (think smeared, or sliding) right or left),
        (Humps: uni- or bi- modal (multi-)   Two humps = two "causes"?)
        Some special shapes:  uniform (p. 40)  && J-shaped (#6 p.50) bell-shaped (Ch 6)
   Center, Spread (roughly now, Ch.5 numerically)
   Outliers,  gaps ? (different groups, sources?)   Look at pulse data.  &&"Lurking  variable"

What do we see?  What can we infer? (Introduction)
    Data source? Lurking variables?
    Variability happens.  Things settle down on average  (Pooled data on colors)
       BUT conclusions are never certain.
    Statistics will give us a language for talking about uncertainty.

Sievers home  Math151-Sp06/Daysp3.htm  3pm 2/3/06
This page belongs to Sally Sievers who is solely responsible for its content. Please see our statement of responsibility.