+ Reply to Thread
Results 1 to 10 of 10

Thread: Standard deviation and variance examples needed

  1. #1
    Points: 2,035, Level: 27
    Level completed: 24%, Points required for next Level: 115

    Posts
    11
    Thanks
    0
    Thanked 0 Times in 0 Posts

    Standard deviation and variance examples needed




    I am trying to calculate SS, variance and standard deviation for the following sample of n = 5 scores: 10, 3, 8, 0, 4.

    Now the sum of squares would be squaring the 10, 3, 8, 0, and 4 and then dividing by 4 (take one out for sampling error) correct?

    Now the mean would be 25/5 (sum of x over n) which is five.

    Taking out variance it would become Sigma(25 (sum of x) - 5 (mean) squared over n-1 (5-1=4)

    So the variance would be 20 squared (400) over four, leaving me with a variance of 100? That doesn't seem quite right to me. Do I have to square root it again to make the s squared into s?

  2. #2
    TS Contributor
    Points: 22,297, Level: 92
    Level completed: 95%, Points required for next Level: 53
    Awards:
    Downloads
    gianmarco's Avatar
    Location
    Italy
    Posts
    1,074
    Thanks
    164
    Thanked 215 Times in 158 Posts
    Hi,
    the variance is equal to the sum of the squared deviation from the mean, divided by the number of observation in your sample. That is, it is the average of the squared deviations from the mean.

    The standard deviation is the square root of the variance.

    Hope this helps.

    Regards
    gm

  3. #3
    Points: 2,035, Level: 27
    Level completed: 24%, Points required for next Level: 115

    Posts
    11
    Thanks
    0
    Thanked 0 Times in 0 Posts
    Quote Originally Posted by gianmarco View Post
    Hi,
    the variance is equal to the sum of the squared deviation from the mean, divided by the number of observation in your sample. That is, it is the average of the squared deviations from the mean.

    The standard deviation is the square root of the variance.

    Hope this helps.

    Regards
    gm
    Thats what I am having problems on, finding the squared deviations. I am not quite sure how to accomplish that. And with the variance, do I square the 100 because as it stands 100 seems an unlikely varianve and its s squared = 100

  4. #4
    Points: 1,938, Level: 26
    Level completed: 38%, Points required for next Level: 62

    Location
    Huntsville, Alabama
    Posts
    46
    Thanks
    0
    Thanked 0 Times in 0 Posts
    You take the difference of each data point from the mean, then square, then sum.

    To get variance, you take the sum squared and divide by n-1 (since you are dealing with a sample).

    Standard deviation is just square root of variance.

  5. #5
    Points: 2,035, Level: 27
    Level completed: 24%, Points required for next Level: 115

    Posts
    11
    Thanks
    0
    Thanked 0 Times in 0 Posts
    Quote Originally Posted by GhostofPythagoras View Post
    You take the difference of each data point from the mean, then square, then sum.

    To get variance, you take the sum squared and divide by n-1 (since you are dealing with a sample).

    Standard deviation is just square root of variance.
    This isn't working out. My variances are ending up in the hundreds and thousands and don't look right. Can you please check my work on the first post and see if I am getting this concept correct?

    BTW, the numbers are the frequency of categorical data.
    To recap let me make up some numbers:

    12, 42, 87, 57, 8

    Squaring = 144, 1764, 7569, 3249, 64

    Adding together is 12790. Using n-1 (which equals 4) I would have a variance of 3,197.5
    Last edited by Deganveran; 02-07-2010 at 09:12 PM.

  6. #6
    Points: 1,938, Level: 26
    Level completed: 38%, Points required for next Level: 62

    Location
    Huntsville, Alabama
    Posts
    46
    Thanks
    0
    Thanked 0 Times in 0 Posts
    Your data set is {10,3,8,0,4}
    Mean = 25/5=5

    Var = {(10-5)^2 + (3-5)^2 + (8-5)^2 + (0-5)^2 + (4-5)^2}/5-1

    (if this was a population instead of a sample, you would divide by 5 instead of 4)

    var = 5^2 + (-2)^2 + 3^2 + (-5)^2 + (-1)^2 / 4

    VAR = 25+4+9+25+1 /4 = 64/4 = 16

    Std Dev = sqrt(16) = 4.

  7. #7
    Points: 2,035, Level: 27
    Level completed: 24%, Points required for next Level: 115

    Posts
    11
    Thanks
    0
    Thanked 0 Times in 0 Posts
    Quote Originally Posted by GhostofPythagoras View Post
    Your data set is {10,3,8,0,4}
    Mean = 25/5=5

    Var = {(10-5)^2 + (3-5)^2 + (8-5)^2 + (0-5)^2 + (4-5)^2}/5-1

    (if this was a population instead of a sample, you would divide by 5 instead of 4)

    var = 5^2 + (-2)^2 + 3^2 + (-5)^2 + (-1)^2 / 4

    VAR = 25+4+9+25+1 /4 = 64/4 = 16

    Std Dev = sqrt(16) = 4.
    Ok I get it now. Sum of squares if squaring the number AFTER you've subtracted the mean from it.

  8. #8
    Points: 1,938, Level: 26
    Level completed: 38%, Points required for next Level: 62

    Location
    Huntsville, Alabama
    Posts
    46
    Thanks
    0
    Thanked 0 Times in 0 Posts
    that is correct.

  9. #9
    TS Contributor
    Points: 18,686, Level: 86
    Level completed: 68%, Points required for next Level: 164

    Posts
    2,802
    Thanks
    9
    Thanked 514 Times in 489 Posts
    Of course you can compute from the statistics ∑x^2 and ∑x:

    S^2 = ∑x^2/(n - 1) - (∑x)^2/[n(n-1)]

    In your case
    ∑x^2 = 10^2 + 3^2 + 8^2 + 0^2 + 4^2 = 100 + 9 + 64 + 0 + 16 = 189
    ∑x = 10 + 3 + 8 + 0 + 4 = 25
    n = 5
    S^2 = 189/(5 - 1) - (25)^2/[5(5-1)] = 189/4 - 125/4 = 16

  10. #10
    Points: 2,101, Level: 27
    Level completed: 68%, Points required for next Level: 49

    Location
    Belgium
    Posts
    87
    Thanks
    0
    Thanked 0 Times in 0 Posts

    Sum of squares is just what it saids... you take the squares, and what is meant by this is actually the difference between an observation and the mean squared. And then you take these squares and sum them up... to get the sum of squares.

    If you devide this by n-1 you get the variance. Sometimes, when you have population data instead of a sample, you devide by n.

    Then, take the square root of this variance to get the standard deviation.

    It's not that hard actually...

+ Reply to Thread

           




Similar Threads

  1. Likert scale data - displaying variance with standard deviation
    By sputnik0110 in forum Psychology Statistics
    Replies: 1
    Last Post: 03-09-2011, 01:00 AM
  2. A simple question about variance/standard deviation
    By Penguin 101 in forum Statistics
    Replies: 9
    Last Post: 01-05-2011, 04:27 PM
  3. Replies: 4
    Last Post: 11-16-2008, 10:03 AM
  4. Expected value, variance, standard deviation
    By i97supratti in forum Probability
    Replies: 3
    Last Post: 12-04-2006, 10:02 PM
  5. Replies: 1
    Last Post: 06-28-2006, 04:52 PM

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts






Advertise on Talk Stats