Business Statistics: Exam I
 Wednesday, January 30, 2013
7:00 pm


You will need a calculator for this exam.  I will provide a list of useful equations and a copy of the normal table.


I. Introduction
A. Using statistics
    1. meaning, context, definition
    2. methodology
    3. descriptive statistics, uncertainty, estimation

B. Reporting statistics
    1. publication bias
        a. "positive" results
        b. sponsor
    2. conclusions may not reflect statistical results
    3. results may not be statistically significant
        - is a confidence level or margin of error reported?
    4. results may not have real world significance
    5. biased samples or biased surveys will give meaningless results
        - is methodology reported?

II. Describing Data

A. Graphs
    1. qualitative variables
        a. pie charts
        b. bar charts
    2. quantitative variables
        a. histograms
        b. boxplots
 
B. Measures of central tendency
    1. mean
    2. median
    3. mode

C. Measures of dispersion
    1. variance
    2. standard deviation
    3. range
    4. quartiles, quintiles, percentiles
    5. coefficient of variation

D. Summarizing distributions
    1. mean, standard deviation
    2. five-number summary
    3. boxplot
 
E. Descriptive Statistics and probability
    1. Chebyshev's theorem
    2.
The Empirical rule (for normal distributions)


III. Relations between variables

    1. scatterplots
    2. covariance
    3. correlation
        a. positive, negative, or zero correlation
        b. strong or weak correlation
    4. regression
 
    5. correlation does not imply causality
        a. X may cause Y
        b. Y may cause X
        c. X and Y may interact
        d. a third  variable may cause both X and Y
        e. the correlation might be spurious

    6. Ordinary Least Squares regression estimates
        a. Y= a+bX
            i. positive, negative, or no relation (slope)
            ii. Predicting value of Y given value of X
            iii. Testing theories
        b. best, linear, unbiased estimate
        c. computer estimation

    7. Regression issues
     a. regression does not show causality
     b. Data collection must be methodologically sound
        i. random sample, representative of population
        ii. avoid bias in survey questions
        iii. results are sensitive to extreme values (outliers)
        iv. model must be correctly specified
            - missing variables
            - spurious relations
 
 IV. Data Collection

A. Samples must be random and representative
            1. sampling design
                a. random sample
                b. systematic sample
                c. stratified sample
                d. cluster sample
            2. response rates
            3. Sample bias
                a. good samples are representative of population
                b. biased samples-- meaningless results
B. Survey Methods
  
  1. write questions carefully
        i. avoid bias
            a. leading questions
            b. confusing questions
            c. over reliance on memory
        ii. group like questions together
    2. test survey
    3. revise


V. Estimation

    1. sample statistics are imperfect estimates of population parameters
    2. variability
        a. sampling error
        b. larger samples reduce variability
    4. standard error
            a. sample mean
            b. proportion
    5. Working with inferential statistics
            a. interval estimates may be preferable
            b. issues of statistical significance  
            c. hypothesis testing

VI. Probability distributions
1. overview.
2. the normal distribution
            a. characteristics
            b. the standard normal
                        i. transforming to standard normal form
                        ii. Using the normal tables

 



Business Statistics class page

Chuck Stull's homepage

Department of Economics

Kalamazoo College Homepage