Saturday, March 21, 2020

Sampling Distributions, Central Limit Theorem and Confidence Intervals

Comic of the Day
Tip: The things themselves are all right, so who cares?


Announcements and Deadlines
Labs 6, 7, 8 and 10 due Thu Apr 16
Excel Final Exam on Thu Apr 16
Chapter 6-8 assignment (Moodle) due Thu Apr 23
Theory Exam (Chapter 6-8) is combined with Chapter 12 exam and will be on Thu Apr 23


This Week

1. Sampling Distributions and Central Limit Theorem

Central Limit Theorem Simulator

Set n=25 and run the simulator several hundred thousand times. 
For any population distribution (normal, uniform, skewed, custom), the distribution of the sample mean, is normally distributed with the following properties:


  • the mean of the sample means approaches the population mean





  • the standard deviation of the sample means is less than the population standard deviation (it actually follows the formula below)

  • Lab 8 - Sampling Distributions
    Lab 8 shows how a sample mean, x̄ (n=16 or 100) is usually close to a population mean, µ (N=10,000), and gets closer as n increases.  You will use a VLOOKUP function to randomly sample from your payroll data.

    3. Confidence Intervals
    "A Mainstreet Technologies poll ... suggests Albertans are highly resistant to any tax hikes. Only 15 per cent of survey respondents say they’d favour increased taxes. Spending cuts are the preferred solution of 43 per cent."
    "...automated phone survey of 3,184 Albertans."
    "The Mainstreet Technologies survey is considered accurate within plus or minus 1.7 percentage points, 19 times out of 20."

    Confidence Interval for Population Proportion, p

    Confidence Interval for Population Mean, µ (σ known)

    Confidence Interval for Population Mean (σ unknown)

    For More Information
    Confidence Intervals (videos go beyond the content of this course)
    Confidence Intervals (small sample size)

    Next Week - Linear Equations and Graphs, Lab 10 - Linear Regression
    Pre-read notes on Moodle

    No comments:

    Post a Comment