01stats

8/11/2019 01stats

1/25

The Benefits

and Risks ofChapter 1

s ng tat st cs

Copyright 2005 Brooks/Cole, a division of Thomson Learning, Inc.

8/11/2019 01stats

2/25

Th h i n 1

A recent news a er article concluded

thatsmoking marijuana at least three

in college.How do you think the researchers came

to this conclusion? Do ou believe it?

Is there a more reasonable conclusion?

Copyright 2005 Brooks/Cole, a division of Thomson Learning, Inc. 2

8/11/2019 01stats

3/25

Th h i n 2Theor : On avera e, men have lower

resting pulse rates than women do.

or disprove that? Would it be sufficient

of each sex? Two members of each sex?What information about mens andwomens pulse rates would help youdecide how man eo le to measure?


8/11/2019 01stats

4/25

Th h i n

Su ose ou were to learn that the lar e

state university in a particular state

went on to become millionaires than any of

.

Would that be a fair comparison?

ow s ou t e num ers e presentein order to make it a fair comparison?


8/11/2019 01stats

5/25

Th h i n 4In its March 35, 1995 issue, USA Weekend

magazine asked readers to return a surveywith a variety of questions about sex andviolence on television. Of 65,142 readers

who responded, 97% were very orsomew at concerne a out v o ence on(USA Weekend, 24 June 1995, p. 5).

about 97% of U.S. citizens are concernedabout violence on TV? Wh or wh not?


8/11/2019 01stats

6/25

1 1 i i

procedures and principles for

gaining and analyzing

information in order to helpeople make decisions when

ace w uncer a n y.


8/11/2019 01stats

7/25

Statistics in Your Own LifeWhich route to go to school,

class, work?Tr each a few timesThen choosebest one.

Sampling routes and comparing them

information to make a decision.


8/11/2019 01stats

8/25

Case Study 1.1: Heart or

ypo a amus

Hypothalamus plays role in emotion, yet most

people think about the heart. Psychologist LeeSalk studied role of heart in human relationships.

Observations:

,by the heart, on 40 out of 42 occasions.

of giving birth), 237 or 83% held babies


8/11/2019 01stats

9/25


ypo a amus

More Observations:

Out o 438 s oppers eav ng store w t onebag, exactly half carried bag on the left.

Denta pat ents as e to o a 5- nc a

while teeth worked on. Substantially more

Salks conjecture:

.

Must be survival value to having a newborn

placed close to sound of mothers heartbeat.


8/11/2019 01stats

10/25


ypo a amus

Controlled Experiment:

Baby nursery at NY City hospital. 1st rou of newborns: Continuous sound

of human heartbeat over loudspeaker.

2ndgroup of newborns: no sound played.Response: weight gain (or loss) after 4 days.

u :Infants treated to heartbeat sound gained


more we g os ess an cr e ess.

8/11/2019 01stats

11/25

1.2 Detecting Patterns

and Relationships

Obvious or Not? Average man taller than average woman.

Taking aspirin helps prevent heart attacks.

More likel to bu blue eans in certain monthsthan others.

Men have lower restin ulse rates than women.

Listening to Mozart improves spatial reasoning.


8/11/2019 01stats

12/25

Statistical StudiesTo conduct a statistical study

1. Get a representative sample.

,

2. Get a large enough sample.

3. Decide whether the stud shouldbe an observational study or

a randomized ex eriment.

Example: Assess if men have lower


.

8/11/2019 01stats

13/25

1. Get a representative sample.

Sample = those actually studied

=

sample was chosen

esearc ers wan o ex en resu s

beyond just the sample.For pulse rates:

Convenient sample = members of your class.

Not valid, e.g., if entire track team in class.Couldnt extend to other age groups.


8/11/2019 01stats

14/25

2. Get a large enough sample.

How large is large enough?

.

More variable within each grouplar er sam le needed to detect a difference.

For pulse rates: a men a ra e o an a women a ra e

of 75, then just need afew to see the difference.

rates ranged from 52 to 82, then would take

more measurements to see a difference.


8/11/2019 01stats

15/25

3. An observational Study

or Randomized Experiment?

Observational Study: merely observe thingsa out our samp e.

Randomized Experiment: randomly assignartici ants to one of various treatment

groups.

Sufficient to measure or observe both pulse

rates and ender of eo le in sam le.

To learn if aspirin helps prevent heart attacks,

a randomized experiment would be better.


8/11/2019 01stats

16/25

Case Study 1.2: Does Aspirin Prevent

eart ttac s

Physicians Health Study (1988)

5-year randomized experiment22,071 male h sicians 40 to 84 ears old .

Group 1: took ordinaryaspirin tablet

every other day. Group 2: tookplacebo (looked like aspirin

but no active ingredients).

Subjects did not know which group they


.

8/11/2019 01stats

17/25

.Heart Attacks?

C o n d it i o n H ea r t A t ta c k N o H ea r t A t t a c k A t ta c k s p er 1 00 0A sp irin 1 0 4 1 0 ,93 3 9 .4 2P lacebo 189 10 ,845 17 .13

Results:

Heart attack rate in as irin rou onl 55% of ratein placebo group.

Randomization other factors, such as amount of

exerc se, expecte to e s m ar or ot groups. Subjects were male physicians. Results may not


, .

8/11/2019 01stats

18/25

1.3 Dont Be Deceived By

Improper Use of Statistics

xamp e

Business Magazine sent 2000 questionnaires torea ers a ou epu can pr mary: rys erpresident Lee Iacocca vs. VP George Bush.

,Iacocca would beat Bush 54% to 47%.

people likely to vote: those who feel stronglyabout an issue most likely to respond.


8/11/2019 01stats

19/25


Example 2

U.S. EPA study led to report that New Jerseyranked 22ndin its release of toxic chemicals.

Problem: wrong units ranking was based

on total pounds released. Using appropriate rate of pounds releasedper

s uare mile, New Jerse ranked 4th on the list.

Figures were misleading until adjusted for size.


8/11/2019 01stats

20/25


Example 3

Article headline:Stud : Smokin Ma Lower Kids I s

Headline implication: a causal connection.

study (since experiment would be unethical).

ora : an t ma e causa connect onsfrom observational studies.


8/11/2019 01stats

21/25

Improper Use of StatisticsNew study confirms too much pot impairs brain

More evidence that chronic marijuana smoking impairs mental

ability: Researchers at the University of Iowa College of Medicine

say a test shows those who smoke seven or more marijuana joints

per week had lower math, verbal and memory scores than non-marijuana users. Scores were particularly reduced when marijuana

Example 4

users e a o n s smo e n e r ungs or onger per o s. SanFrancisco Examiner, 13 March 1993, p. D-1.

Headline implication: a causal connection.

Can concludethere is a relationshipbut


.

8/11/2019 01stats

22/25

Case Study 1.3: A Mistaken Accusation

o eat ng

Two students accused of cheating on multiplex ue o w sper ng ur ng one

part and wrong answers often matched each other. Licensin board estimated odds of havin answers

as close as these two were < 1 in 10,000.

At trial, statistician showed agreement in this examwas no g er t an n ot er exams.

Key: two students were husband and wife, both.

they were just guessing. Instead, common mistakeswere often the right answers for India background.


Whispering had to do with status of sick child.

8/11/2019 01stats

23/25

1.4 Summary and Conclusions

using statistical methods.

You need to know how data were collected,, .

Over the next several chapters, you will

ecome an e uca e consumer o s a s cainformation.


8/11/2019 01stats

24/25

Tutorial Questions

,

such as movies, sports events and concerts,

cousins to survive the next eight to nine

years, even w en e uca on an ncome are

taken into account, according to a survey by

t e n vers ty o mea n we en(American Health, April 1997, p.20).


8/11/2019 01stats

25/25

Tutorial Questions

.

randomized experiment? Explain.b. On the basis of the stud that was conducted can

we conclude that attending cultural events causes

people to be likely to live longer? Explain.c. The article continued No ones sure how Mel

Gibson and Mozart help health, but the activities

may enhance immunity or coping skills.Comment on the validity of this statement.


Documents

01stats