27
1 Chapter 1: How do we get “good” data?

Chapter 1: How do we get “good” data?

  • Upload
    ronat

  • View
    33

  • Download
    1

Embed Size (px)

DESCRIPTION

Chapter 1: How do we get “good” data?. What does the word “statistics” mean to you?. Definition Applications Where you’ve seen statistics before Your feelings about statistics …. Course Layout. More conceptual than computational I will give reading assignments very often. - PowerPoint PPT Presentation

Citation preview

Page 1: Chapter 1: How do we get “good” data?

1

Chapter 1: How do we get “good” data?

Page 2: Chapter 1: How do we get “good” data?

2

What does the word “statistics”mean to you?

• Definition

• Applications

• Where you’ve seen statistics before

• Your feelings about statistics …

Page 3: Chapter 1: How do we get “good” data?

3

Course Layout

• More conceptual than computational– I will give reading assignments very often.

• More frequent smaller quizzes• Book breakdown:– I. Producing Data

– II. Organizing Data

– III. Chance

– IV. Inference

Page 4: Chapter 1: How do we get “good” data?

4

Questionnaire

Page 5: Chapter 1: How do we get “good” data?

5

Opening Day Questionnaire

• In groups of 3, compile the data (for #2, #7) and prepare a short report of a couple of things.

– Done on whiteboards.

– Graphs, tables, statistics, etc.

– Color!

Page 6: Chapter 1: How do we get “good” data?

6

Questionnaire

• What are the individuals? Variables?

– Definitions, p. 5

• Type of study?

– Observational? (p. 9)

– Experiment? (p. 16)

Page 7: Chapter 1: How do we get “good” data?

7

Homework

• Reading, pp. 3-17

• Prepare a dotplot for questionnaire item #9.

– See Activity 1.1, p. 4.

• Exercises 1.1 and 1.4, pp. 7-8

Page 8: Chapter 1: How do we get “good” data?

8

Comparing Observational Studiesand Experiments

• Definitions, p. 9 and p. 16

• Give two examples of each.

Page 9: Chapter 1: How do we get “good” data?

9

Populations and Samples (p. 10)

• Population: The whole thing

• Sample: A subset of the whole thing

– Statistics is usually concerned with taking a sample to infer something about the population.

• Census (p. 13): Entire population is included in the sample (or at least there is an attempt to do so).

Page 10: Chapter 1: How do we get “good” data?

10

Exercises

• 1.8, p. 13

• 1.12, p. 17

Page 11: Chapter 1: How do we get “good” data?

11

Homework

• Read: Statistics in Summary, p. 20• 1.15, p. 18• 1.20 and 1.23, p. 21• Read: pp. 22-35• Section 1.1 quiz on Thursday• Extra credit opportunity:– Application 1.1, p. 19

– Due on or before 1.18.09 (Monday)

Page 12: Chapter 1: How do we get “good” data?

12

Section 1.2: Measuring

• We must have an operational definition of the construct we want to measure.

– For example, it’s one thing to say we want to measure intelligence (the construct), but it is quite another to actually measure it (operational definitions).

• Valid measure: p. 28

Page 13: Chapter 1: How do we get “good” data?

13

Valid Measurements for …

• Physical fitness

• Happiness

• “Well-educated”

• Student “readiness” for college

Page 14: Chapter 1: How do we get “good” data?

14

USDA Statement on Laura Lynn 2% Milk (which does not contain rBGH growth hormone)

• “Milk from a cow supplemented with rbGH is not different from that of a non-supplemented cow.”

• See sidebar, p. 33

– “The Great One”

Page 15: Chapter 1: How do we get “good” data?

15

Predictive Validity (p. 31)

• Application 1.2A, p. 32

• Excel file: Predictive validity for SAT at Rice University

• Employment law:– http://www.employment-testing.com/validity.htm– Sonia Sotomayor article in New York (hiring practices

for fire fighters): http://www.newyorker.com/reporting/2010/01/11/100111fa_fact_collins?currentPage=all

Page 16: Chapter 1: How do we get “good” data?

16

Measurement Definitions

• p. 24:

– measure, instrument, units, variable

• Exercise 1.24, p. 27

Page 17: Chapter 1: How do we get “good” data?

17

Homework

• Look over examples 1.14 and 1.15, p. 30

• Exercises:

– 1.31 and 1.32, p. 33

– 1.34, p. 34

• Reading: pp. 34-42

Page 18: Chapter 1: How do we get “good” data?

18

Measurement Validity

• We’ve spoken about the need for a measurement to be valid.– Definition, p. 28

• Ways we establish evidence of validity:– Predictive validity (e.g., SAT vs. college GPA)– Face validity: Have a panel of experts (SME) study our

instrument for measuring.• There are statistics for measuring this (dissertation, p. 41)

– Statistical methods• Correlations with other similar measurements• Use as independent variable in designed experiments

Page 19: Chapter 1: How do we get “good” data?

19

Measurement Reliability (p. 35)

• In addition to using valid measurements, our measurements must be reliable.

– Reliable=repeatable results

• Ways to establish evidence of reliability:

– Test-retest

– Parallel tests

– Statistical methods, including internal consistency evaluations.

Page 20: Chapter 1: How do we get “good” data?

20

Bias (p. 35)

• Systematically overstates or understates the true value of a property.

Page 21: Chapter 1: How do we get “good” data?

21

Bias and Reliability

Page 22: Chapter 1: How do we get “good” data?

22

Scales Example

Page 23: Chapter 1: How do we get “good” data?

23

Practice

• See Example 1.17, p. 35

• Exercises:

– 1.35, p. 39

– 1.42, p. 42

– 1.44, p. 43

– 1.48, p. 44

Page 24: Chapter 1: How do we get “good” data?

24

More practice, section 1.2

• Exercises, pp. 39-40:

– 1.37,1.38,1.39,1.41

• Section 1.2 quiz tomorrow (Tuesday)

Page 25: Chapter 1: How do we get “good” data?

25

Section 1.3: Do the numbers make sense?

• What they did not tell us … numbers have a context

– p. 46

• Are the numbers plausible?

– p. 49

• Are the numbers too good to be true?

– p. 50

– Fake data? Too precise?

• Is the arithmetic right?

– p. 51

• Is there a hidden agenda?

– p. 53

Page 26: Chapter 1: How do we get “good” data?

26

Section 1.3 problems

• pp. 55-58:

– 1.55, 1.59, 1.62, 1.64

Page 27: Chapter 1: How do we get “good” data?

27

Chapter 1 Review Exercises

• pp. 59-62:

– 1.71, 1.73, 1.75, 1.79