Descriptive Statistics - PowerPoint PPT Presentation

1 / 20
About This Presentation
Title:

Descriptive Statistics

Description:

Shoes. Foot. Fit. 5. Descriptive Statistics ... Use Excel, Minitab, SAS, etc. 15. 16. Histogram of scores in Quiz 1. Descriptive Statistics ... – PowerPoint PPT presentation

Number of Views:46
Avg rating:3.0/5.0
Slides: 21
Provided by: Ni74
Category:

less

Transcript and Presenter's Notes

Title: Descriptive Statistics


1
Descriptive Statistics
  • 01/24/2006

2
Review
  • Probability Theories
  • Normal Distribution Statistical Models
  • Calculate normal probabilities
  • Central limit theorem
  • Other distributions in Normal family

3
What do we observe in reality?
  • Hedge Funds Investment Example
  • If we invest x in mutual fund I, its worth
    after 1 year is distributed X_I N(1.05 x,
    0.0002 x2)
  • Weight (Y) distribution of students in 2028B
  • Y N(160, 100)

4
Relationship of Data and Model
  • Shoes
  • Foot
  • Fit
  • Models Normal, Binomial, t
  • Data (x1, x2, )
  • Fit

5
Descriptive Statistics
  • Part I. ? X is a random variable (w/probability
    distribution.)
  • We assume knowledge about X
  • distribution (parameters)
  • mean, variance, etc
  • Part II. ? Study of sample data (X1, ,Xn)
  • We dont know all the parameters
  • if we did, why get data?
  • use data to estimate the unknown parameters
  • use data to make decisions about parameters
  • Connection Assume data (X1, ,Xn) are
    realizations of a random variable. We
    characterize its distribution by estimating
    parameters

6
Probability distribution f(x)
Experimentation
Probability theory
Data set x1, x2, xn
Data analysis
Statistical Inference
Estimate properties of f(x)
7
(No Transcript)
8
Data exploration and presentation
  • Categorical data (Nominal data)
  • Numerical data
  • --integers
  • --real numbers
  • Charts
  • Collect observations into groups to construct
    frequency histogram. Often, this is the BEST way
    to communicate your findings from a set of data
    (People understand histograms)

9
Categorical data
10
Bar Chart
11
Misleading Bar Chart
12
(No Transcript)
13
Pie Chart
14
Histogram
  • We can (generally) rely on computers to generate
    histograms
  • Keep bin widths equal
  • Choose width that summarizes the data best
  • Use Excel, Minitab, SAS, etc

15
(No Transcript)
16
Histogram of scores in Quiz 1
Descriptive Statistics Mean 8.120 Standard
Deviation 1.740
17
(No Transcript)
18
Conveying Information Graphically
Time Series
19
(No Transcript)
20
  • Quiz 2 this Thursday
  • Homework 1 is due on this Thursday
Write a Comment
User Comments (0)
About PowerShow.com