Communicating Quantitative Information - PowerPoint PPT Presentation

About This Presentation
Title:

Communicating Quantitative Information

Description:

Communicating Quantitative Information Midterms. Compound interest. Should you guess on . Out-of-wedlock births (census data) Jury selection. – PowerPoint PPT presentation

Number of Views:81
Avg rating:3.0/5.0
Slides: 35
Provided by: Jeani169
Category:

less

Transcript and Presenter's Notes

Title: Communicating Quantitative Information


1
Communicating Quantitative Information
  • Midterms. Compound interest. Should you guess on
    .
  • Out-of-wedlock births (census data)
  • Jury selection.
  • Homework Complete Project I (turn in proof-read
    paper and/or 1 pager and be prepared to present)

2
Election
  • Latest www.fivethirtyeight.com
  • Especially good on models of electorate,
    confidence interval, spread
  • VOTE
  • Good thing to do for multiple reasons
  • College loans issue, get the habit, other
  • Youth vote may be better than expectations
  • Extra credit
  • Bring picture or some other evidence

3
Reprise Excel
  • Frequency calculation sort data into bins, count
    the number in each bin
  • Named cells
  • Various functions
  • median, average each take an array as operand
  • power takes number and exponent
  • Sort
  • Note we didn't do multiple columns
  • Carry along or use for sorting

4
Preview Excel
  • (where and when we do this will depend on time
    required for presentations)
  • Compound interest
  • Aka time value of money
  • 1000 grows to 1050 grows to (10501.05)

1000 1050 1103 1158 1216 1276 1340 1407
5
Preview
  • Mortgage on property (house)
  • Borrow money according to some agreed upon
    conditions
  • Fixed interest rate, fixed amount of time (30
    years)
  • Floating interest rate, rate changes at fixed
    times based on other rates
  • Fixed rate for fixed time, then so-called balloon
    rate
  • For all these, the bank has a lien on the house
    foreclosure on the house means that bank takes
    over because payments were not made.

6
Sub prime mortgage
  • People who were considered high risk (sub prime
    borrowers) were given different conditions
  • Higher rate
  • Lower initial rate and then higher rate
  • interest only and then so-called balloon payment
  • Many of these loans have not been paid
  • At some point, the loaners decide to give up
    "Loans on the books" ? Bad debts
  • In place of a revenue stream, instead have
    property that can be sold, but probably at lower
    prices and at a cost in staffing

7
Typical case (old days) Home equity/second
mortgage loan
  • Person made substantial down payment (20) Say
    20,000 of 100,000 home.
  • Prices / value of house (apparently) went up.
    Suppose 100,000 to 150,000
  • As a consequence, owner still owed the bank, but
    had increased equity. Say 20,000 50,000 5000
    (paid off) 75000
  • Many people went to bank and borrowed against
    this 75000! Used it for other bills.
  • "using home as ATM"

8
Typical sales pitch
  • For original loans and for home equity/second
    mortgages
  • Houses are going up, so if the loan gets to be
    too much, you can sell!
  • BUT what happens when smaller or zero down
    payment AND/OR housing prices go down?

9
Trade in mortgages
  • A bank (loaner) can sell a mortgage to someone
    else.
  • I'm giving you the IOU from X. We agree on a
    price. He now pays you.
  • Securitized mortgages Banks can and did bundle
    mortgages and sell them to third parties who sold
    shares in the mortgages to investors.
  • This led to demand for more mortgages and less
    oversight/interest in identifying good customers
    and working with people in trouble.
  • Who actually owns the mortgage? Who can
    foreclose?

10
Next computer lab class
  • Do original and second mortgage examples.
  • pmt (interest,numberpayments, principal)
  • computes the payment to make, assuming
    numberpayments constant payments, to pay back a
    loan of principal amount at interest
  • Each payment pays off some of the principal plus
    the interest owed at that point.
  • over time, more of the payment is towards the
    principal and less covering the interest.

11
Should you guess
  • On my tests, yes, because I give partial credit
    and I don't count off.
  • But what about standardized tests typical
    formula ispenalty of ¼ for wrong answers
  • Should you guess?

12
Answer
  • Assume 5 multiple choice answers
  • If you have no idea
  • What is the expected value of a guess?1/5 of the
    time the value is 1 4/5 of the time the value is
    -1/4
  • Total is 1/5 1 4/5 times (-1/4)
  • Total is 0! On average, no benefit, but no
    penalty from guessing.

13
Answer, continued
  • What if you can remove one of the 5 answers, then
    expected value is
  • ¼ of the time value 1
  • ¾ of the time the value is -1/4So total is ¼1
    ¾ (-1/4) ¼ 4/16 ¾ (-1/4) -3/16Total
    is 4/16 3/16 1/16. This is positive, so it
    is worth guessing!

14
My conclusion
  • You should guess if
  • guessing does not take too long
  • or get you too nervous
  • you can remove 1 or 2 choices

15
Test strategy
  • Study
  • Practice taking practice tests, if available
  • If you can rule out some of the answers AND IT
    DOES NOT TAKE TOO LONG, guess from remaining
    answers.

16
Out-of-wedlock births
  • Fairly regular news story
  • One or more populations have increases in
    out-of-wedlock births
  • As previous situations
  • Definitions matter
  • Must make note of issues and terms such as
    percentages, rates, changing underlying
    population, absolute vs rates, rates of change,
    time interval under study

17
Challenge
  • Let's pose some questions and see if we can find
    answers.

18
http//www.cdc.gov/nchs/fastats/unmarry.htm
  • Unmarried Childbearing
  • (Data are for U.S. in 2007)
  • Number of live births to unmarried women
    1,715,047
  • Birth rate for unmarried women 52.3 births per
    1,000    unmarried women aged 15-44 years
  • Percent of all births to unmarried women 39.7 
  • Source  Births, Final Data for 2007

19
Other source (story)
  • http//www.acf.hhs.gov/programs/ofa/annualreport5/
    chap08.htm
  • "Bonus to Reward Decrease in Illegitimacy Ratio"
    to as many as five States (and three Territories,
    if eligible) that achieve the largest decrease in
    out-of-wedlock births without experiencing an
    increase in their abortion rates above 1995
    levels.

20
Murphy Brown story
  • TV character!!! choosing to have child as a
    single mom. Criticized by VP Dan Quayle
  • News story Murphy Brown was accurate
    Out-of-wedlock births by professional,
    well-off, non-minority women increasing at
    greater rate than other groups.
  • Be careful when comparing rates of change
  • If 1 goes to 2, this is a doubling,100
    increase
  • If 45 goes to 50, this is increase of 11.1

21
Jury selection
  • From Cartoon Guide to Statistics.
  • Based on real stories (may be combination)

22
Notice
  • Jury panel
  • intended to be random selection of
  • voting list
  • driver's license?
  • reachable at address
  • versus
  • Jury
  • dismissed for cause
  • each side gets some number preemptory challenges
    chances to dismiss
  • Attorneys use the jury selection process to
  • find out about jury members
  • begin to make their case

23
Jury panel
  • Panel called up from voting lists (maybe other
    lists as well)
  • Juries are chosen from the panels.
  • May be that previous (though not recent)
    experience on jury makes it more likely that you
    will be called
  • I have served on local (Mt. Kisco), county (3
    times), federal (alternate), and federal grand
    jury
  • Rules changed so not as easy to avoid service
  • I RECOMMEND IT!

24
Case in textbook
  • Area was 50 African-American, 50 White in
    population.
  • Panel consisted of 80 people, 4 African-American.
  • What is probability of such a panel if selection
    was totally random?
  • (What is the probability of flipping a fair coin
    80 times and only getting 4 heads?)

25
Calculation
  • The TOTAL number of different strings of
    heads/tails is 2 raised to the 80th power
  • 2 2 2 (80 times)
  • How many of these have exactly 4 heads?
  • Heads can be anywhere in the string of 80.
  • Let's take a detour and say the panel is the size
    of the people in the class.

26
This class
  • Random means at any time, any student can be
    picked
  • What are the number of ways to pick exactly 3
    males when selecting a panel of 6?
  • M M M F F F, M M F M FF, etc.
  • The answer is the same as choosing 3 places
    among the 6
  • 6 5 4
  • Were not donethis is an overcount because it
    counts M1 M2 M3 F F F and M2 M1 M3 F F F as
    distinct, so need to divide by the number of ways
    you can shuffle 3
  • (654) / (321)
  • (Note this class is NOT half male and half
    female.)

27
Back to problem
  • Write down all the patterns 80 long of Heads and
    Tails (or some other designation, e.g., W and A
  • How many have exactly 4 As and all the rest W?

28
Back to the problem
  • 80 positions. Choosing which positions are
    'Heads'
  • Pick 4 from the 80 80797877
  • Divide by 4321, because you don't care about
    order of picking
  • The total number is something less than 1/280
    because once someone is chosen, then that person
    is removed from the pool, so the ratio of White
    to African-American changes slightly. However,
    given a large population (gtgt 80), this is a good
    estimate.

29
Calculation
  • ((80797877)/(4321))/(280)
  • Using Excel as is ORbinomdist(number,trials,prob
    ability,cumulative)
  • Answer is 1.30825E-18
  • This is 0.00000000000000000130825
  • Cartoon Book notes that this is less than the
    chances of getting 3 royal flushes in a row in
    poker.
  • Judge ruled that it could not have been random
    and the court had to change.

30
Alito story
  • Riley v. Taylor
  • Appeal found in favor of defense based on
    statistical analysis on absence of
    African-Americans on jury
  • mis-use of pre-emptory challenges
  • Alito dissented, saying only 10 of Americans are
    left-handed and yet 5 of the last 6 elected
    presidents are left-handed
  • Bush, Clinton, Bush Senior, Reagon, Carter, Ford

31
Counter-arguments
  • Independent of statistics, there is a history of
    racial bias
  • Facts are wrong
  • Ford not elected. Nixon was right-handed
  • Reagan and Bush, Jr also right handed
  • So, of the last 6 elected, only 2 left-handed
  • Wrong to stop at arbitrary point of 43
    presidents (elected and not-elected), only 4
    left-handed
  • Note left-handedness may not be reported
    accurately
  • What do you think?
  • http//chance.dartmouth.edu/chancewiki/index.php/C
    hance_News_9

32
Sub-groups
  • Picking the last 6 presidents
  • You may pick the ones that suit your argument by
    deciding when to stop.
  • Recent Women's Health Initiative study,
    discussion (disagreements) on findings of
    sub-groups.

33
Homework
  • Postings
  • Complete Project 1
  • essay (proofread!), charts and diagrams (at least
    1)
  • Proof read manually and
  • Use spell and grammar check
  • Ignore, ignore all when appropriate
  • presentation (include 1 pager!!!)
  • Proof read 1 pager AND charts

34
Themes
  • What are the definitions?
  • Is there any specialized language (jargon) here?
  • Confidence level
  • out of what?
  • What is the universe? What is the domain of
    discussion?
  • What is the context
  • Time, space (geography)
  • What else? If not this, what
  • People who get medical treatment are generally
    sick, so need to be compared to other sick
    people, not well-people!
  • What else?
  • What factors are not mentioned?
  • Do things add up?
Write a Comment
User Comments (0)
About PowerShow.com