Other's On-Line Resources (Chapter 8)

 

Statistical Power
  • Description:
    By using this interactive resource, you'll be able to (1) see a wonderful picture that illustrates graphically the concept of statistical power and (2) change one or more of four factors (e.g., sample size) snd then observe what happens to power as a consequence of your change(s).

  • What to Do:
    1. Click on the colored title of this on-line resource: "Statistical Power."
    2. Scroll down to the bottom of the screen that pops up.
    3. Click on the "Show it" button below the large gray box.
    4. Enter this information in the various windows below the gray box: True mean = 110, Hyp. mean = 100, sigma = 15, and n = 25.
    5. Click again on the "Show it" button, and then look inside the gray box at (a) the picture and (b) the power value presented beneath the picture.
    6. Increase n to 50 and then click on "Show it" to see what happens.
    7. Change other information in the input boxes and then click on "Show it" to see what happens.

  • Sky Huck's Puzzle Question:
    The power value shown inside the large gray box is influenced by 4 decisions you make concerning: (a) the difference between the true and hypothesized means, (b) sigma, (c) n, and (d) whether the test is one-tailed or two-tailed. Which of these 4 things have a direct relationship with power, and which have an indirect relationship?
     
Practical vs. Statistical Significance (A)
  • Description:
    By using this interactive on-line resource, you'll be able to see how statistically significant results can come about even though the null hypothesis is just slightly false.

  • What to Do:
    1. Click on the colored title of this on-line resource: "Practical vs. Statistical Significance (A)."
    2. After clicking on "Begin," click "Simulate" in the next screen you see. By doing this, you'll produce data that will be analyzed to see if a statistically significant difference exists between the mean of a sample drawn from Population A and the mean of a sample drawn from Population B. Whether your result is "Significant" or "Not significant" will be shown in this screen's upper right-hand corner.
    3. Now, change the mean of Population B from 15 to 10.001. By doing this, you've made the 2 population means different, but different by only a VERY TINY amount. In other words, the null hypothesis (of equal population means) is false, but it's false to only a very trivial degree.
    4. Click the "Reset" button that's above the gray box.
    5. In the top row of windows, change "n" from 8 to 30, change the mean for Population B from 15 to 10.001, change the "Population rho" from 0.5 to 0.999, and change "Population sd" from 5 to .01.
    6. Click "Simulate" to see if the sample means from the nearly identical populations are significantly different. If so, you've just proven to yourself that it's possible for the hypothesis testing procedure to yield a statistically significant result when the null hypothesis is "off" by the tiniest of margins. What you produced was a statistically significant result . . . but it had NO practical significance. In other words, what you created was the illusion of a "mountain" when, in reality, what existed was only a tiny little molehill.

  • Sky Huck's Puzzle Question:
    Repeat Steps #3 and #4, but this time make the mean of Population B equal to 10. Before clicking on "Simulate 5000" to replicate your study 5,000 times, make a guess as to what the "Percent significant" number will be. After making your guess, click on "Simulate 5000" to see what happens.
     
Practical vs. Statistical Significance (B)
  • Description:
    Although this site is non-interactive, it contains a simple example designed to show that "statistical significance" does not necessarily imply "practical significance."

  • What to Do:
    1. Click on the colored title of this on-line resource: "Practical vs. Statistical Significance (B)."
    2. Read carefully what's said on each screen, moving from one screen to another by clicking on the "Next" buttons.
    3. You'll come across several words that are links to brief but helpful explanations of what a term means. Don't pass up the opportunity to gain a handle on these terms by using the available links.

  • Sky Huck's Puzzle Question:
    Look at the picture displayed on the 2nd of the 4 pages of this on-line resource. Are you looking at confidence intervals?
     

Copyright © 2003 Schuyler W. Huck
All rights reserved.

| Book Info | Author Info |

Site URL: www.readingstats.com
Top | Site Map
Site Design: John W. Taylor V