USING CALCULATORS AND COMPUTERS IN STATISTICS

1
USING CALCULATORS AND COMPUTERS IN STATISTICS

• Laura J. Niland
• MacArthur High School
• 2923 Bitters
• San Antonio, TX 78217
• 210-650-1100
• e-mail ljniland_at_texas.net
• Joe Ward
• Health Careers High School
• 4646 Hamilton Wolfe
• San Antonio, TX 78229
• 210-433-6575
• e-mail joeward_at_tenet.edu

CAMT98 45th Annual Conference San Antonio,
Texas July 23, 1998
2
PREVIEW OF USING CALCULATORS AND
COMPUTERS IN STATISTICS

• OVERVIEW OF GENERAL APPROACH TO PROBLEM
  ANALYSIS
• FOCUS ON THE COMBINED POWER OF REGRESSION MODELS
• AND COMPUTERS
• TWO-CATEGORY t- TEST FROM MOORE MCCABE -IPS
• USING REGRESSOIN MODELS
• THREE-CATEGORY ANOVA FROM MOORE MCCABE
• USING REGRESSION MODELS
• MANATEE REGRESSION MODEL FROM MOORE MCCABE

3

CONNECTING ANALYSIS OF VARIANCE(ANOVA)
ANDREGRESSION MODELS

Combining Regression Models and Statistical
Software Students can
-- Do useful, meaningful analyses AFTER the
learning experience that they could not do
BEFORE.
-- Analyze many different-appearing data analysis
procedures with one general approach.

-- Reduce the risk of unknowingly obtaining
irrelevant output from statistical software.

-- More easily and correctly specify
computational requirements to the computer.


-- Simplify communicating results of the
analyses.
4
THE BIG PICTURE
Prediction
Uncertainty
Optimization
Modeling

Problem Solving
Through Data Analysis
Mathematics Curriculum
Linking
Linking
Disciplines
Different
Different
Statistics
t-tests ANOVA Regression
Biological, Physical, Social
Sciences
Engineering
5

ENRICHING A FIRST STATISTICS COURSE USING
PREDICTION MODELS AND STATISTICAL SOFTWARE

1
3
Identify real-world problems
Learn to use a computer
2
4
Translate questions of interest into
mathematical models
Learn to use statistical software
5
Use statistical software for computational
requirements
6
7
Interpret results from statistical software
Translate to real-world actions
6
Translate into Statistical Models
General form of PREDICTION MODELS REGRESSION
MODELS LINEAR MODELS
Y(Response) c1X1 c2X2 c3X3 . . .
c(last) X(last) E(rrors)
Let P(redictions) c1X1 c2X2 c3X3 . .
. c(last) X(last)
then Y(Response) P(redictions) E(rrors)
As expressed throughout Introduction to the
Practice of Statistics by Moore McCabe DATA
FIT RESIDUAL
7
(No Transcript)
8
(No Transcript)
9
PREDICTION is the Name of the Game

• Predict BP Change Knowing TREATMENT (Calcium or
  Placebo)

• Predict Reading SCORE Knowing METHOD (Basal,
  Drta or Strat)

• Predict MANATEES Killed Knowing NUMBER of BOATS

10
(No Transcript)
11
(No Transcript)
12
(No Transcript)
13
(No Transcript)
14
(No Transcript)
15
(No Transcript)
16
See the results of this analysis in Example 7.12,
pp. 544-546 in Moore and McCabe INTRODUCTION to
the PRACTICE of STATISTICS Second Edition,
1993 by W.H. Freeman
17
(No Transcript)
18
(No Transcript)
19
(No Transcript)
20
(No Transcript)
21
(No Transcript)
22
See the results of this analysis in Example 10.6,
pp. 725-732 in Moore and McCabe INTRODUCTION to
the PRACTICE of STATISTICS Second Edition,
1993 by W.H. Freeman
23
(No Transcript)
24
(No Transcript)
25
(No Transcript)
26
(No Transcript)
27
(No Transcript)
28
See the results of this analysis in Section 9.1
Exercises, pp. 674-675 in Moore and
McCabe INTRODUCTION to the PRACTICE of
STATISTICS Second Edition, 1993 by W.H. Freeman
29
(No Transcript)
30
(No Transcript)