11-13-2016, 04:22 PM
I'm helping somebody prepare for the DSST statistics exam. I found a website that has something worth reading. Manual Computation of the variance and standard deviation
In math, there are definitional formulas and computational formulas. Definitional formulas are good at explaining a concept. In the case of the definitional formula for the standard deviation, with only a little bit of explanation, it's easy to see that values further from the mean impact the magnitude of the standard deviation more than values close to the mean. That's great to help you conceptually, but not so great for helping you calculate a standard deviation by hand.
Computational formulas are mathematically equivalent to definitional formulas. They are arranged differently to make it easier to compute the answer by hand. Why should you care?
Although I could add a Unicode character for sigma to this post, it might not render properly on your device. I will use E instead of sigma. What if you aren't asked to compute the standard deviation from a bunch of numbers? What if instead you are given values corresponding to parts of a formula? Suppose you are given (E X) squared and (E Xsquared) instead of all of the raw data. Would you know what to do?
Now consider the computational formula for the correlation coefficient. Computing Pearson's Correlation Coefficient (David Lane and Rice University published an excellent free statistics textbook, see Free Statistics Book)
What if you were provided the following
E XY
E X times E Y
E Xsquared
E Ysquared
(E X) squared
(E Y) squared
If you know the formula, you should be able to make a few substitutions and quickly find the answer. Just be sure to know if you should use n or n-1.
Now to state what some may find obvious. If you don't understand what this post is talking about, you're not ready for the exam. It's assumed you've already spent some time studying statistics to understand the post.
Please add feedback about the $5 official DSST practice test for statistics and the $20 Peterson's practice tests. https://www.nelnetsolutions.com/testprep...stname=262 There's also a free practice test DSST Principles of Statistics: Study Guide & Test Prep - Practice Test Questions & Final Exam | Study.com
In math, there are definitional formulas and computational formulas. Definitional formulas are good at explaining a concept. In the case of the definitional formula for the standard deviation, with only a little bit of explanation, it's easy to see that values further from the mean impact the magnitude of the standard deviation more than values close to the mean. That's great to help you conceptually, but not so great for helping you calculate a standard deviation by hand.
Computational formulas are mathematically equivalent to definitional formulas. They are arranged differently to make it easier to compute the answer by hand. Why should you care?
Although I could add a Unicode character for sigma to this post, it might not render properly on your device. I will use E instead of sigma. What if you aren't asked to compute the standard deviation from a bunch of numbers? What if instead you are given values corresponding to parts of a formula? Suppose you are given (E X) squared and (E Xsquared) instead of all of the raw data. Would you know what to do?
Now consider the computational formula for the correlation coefficient. Computing Pearson's Correlation Coefficient (David Lane and Rice University published an excellent free statistics textbook, see Free Statistics Book)
What if you were provided the following
E XY
E X times E Y
E Xsquared
E Ysquared
(E X) squared
(E Y) squared
If you know the formula, you should be able to make a few substitutions and quickly find the answer. Just be sure to know if you should use n or n-1.
Now to state what some may find obvious. If you don't understand what this post is talking about, you're not ready for the exam. It's assumed you've already spent some time studying statistics to understand the post.
Please add feedback about the $5 official DSST practice test for statistics and the $20 Peterson's practice tests. https://www.nelnetsolutions.com/testprep...stname=262 There's also a free practice test DSST Principles of Statistics: Study Guide & Test Prep - Practice Test Questions & Final Exam | Study.com
63 CLEP Sociology
75 CLEP U.S. History II
63 CLEP College Algebra
70 CLEP Analyzing and Interpreting Literature
68 DSST Technical Writing
72 CLEP U.S. History I
77 CLEP College Mathematics
470 DSST Statistics
53 CLEP College Composition
73 CLEP Biology
54 CLEP Chemistry
77 CLEP Information Systems and Computer Applications
75 CLEP U.S. History II
63 CLEP College Algebra
70 CLEP Analyzing and Interpreting Literature
68 DSST Technical Writing
72 CLEP U.S. History I
77 CLEP College Mathematics
470 DSST Statistics
53 CLEP College Composition
73 CLEP Biology
54 CLEP Chemistry
77 CLEP Information Systems and Computer Applications