Page 11: Questions about Scores and Probability

We know something important: scores that are close to the mean are average in two ways

  They are neither very high nor very low.

  They are very frequent.

Scores that are much higher or lower than the mean are rare.

We can transform a distribution of raw scores into a distribution of z scores.  If  the raw score distribution is bell-shaped and symmetrical, we use the properties of the normal distribution to estimate the probability of observing different ranges of raw scores.  Click here (Mark-there is an error in this link)to construct a frequency histogram of both the conscientiousness data and the conscientiousness data transformed into z-scores.

   Click on the statements below which describe the z distribution of transformed conscientiousness scores? 

The z distribution has the same shape as the raw score distribution.
Both the z and the raw score distribution are somewhat negatively skewed.
Both the z and the raw score distribution are somewhat positively skewed.
The z distribution is bell shaped and symmetrical, but not the raw score distribution.


Q. If a raw score distribution is bell shaped and symmetrical, then transforming it into a z distribution will create a bell shaped and symmetrical distribution as well.   Assume that this situation holds, and that a score of 20 in a hypothetical raw-score distribution transforms to a z score of 1.  If you look up the probability of a z score equal to or less than +1.0 in a standard normal table, what is the relation between this probability multiplied by 100 and the percentile rank of the raw score of 20. 

The standard-normal probability * 100 will be larger than the percentile rank.
The percentile rank will be larger than the standard-normal probability * 100.
Within rounding error, the two values should be the same.


Page 12.-Computing probabilities for the conscientiousness data.

Now let's calculate the probability for some raw scores on the conscientiousness scale.   Because the conscientiousness distribution is not quite symmetrical, the probabilities computed using the normal distribution will be slightly different from the percentile ranks computed from the raw score distribution itself.  We will now examine how closely the standard normal probabilities approximate the empirically derived percentile ranks in value.  In order to do this, you will need to open two windows:  one that shows the frequency histogram for the raw score distribution and one that shows the standard normal distribution.

Click here (Mark-there is an error in this link)to create a frequency histogram, with an interval width of 1, for the raw score conscientiousness  distribution.  You can find the percentile rank of each raw score in the distribution by clicking on the bar in the histogram directly to the right of the bar over the targeted raw score or by clicking on the row in the table directly below the row containing the targeted raw score.

Click here to create a standard normal distribution.  Clicking on 'sampling distribution of the mean' under the 'sampling distributions' menu will open a window containing a standard normal distribution.  There are two arrows under the x-axis of the distribution.  Put the arrow at the left-hand end of the distribution at -5.0 (or minus infinity), and move the arrow at the right-hand end of the distribution until it points to the value of z you are interested in (values of z appear in yellow highlighting.  The area under the curve between minus infinity and  z  appears in blue and the probability of observing values of z between minus infinity and z is the numerical value in red at the top of the window.

The conscientiousness distribution has a mean of 24.10 and a standard deviation of 3.17.  To answer the questions below, use pencil and paper to compute the value of z for the raw score queried.  Find the probability in the standard normal distribution and multiply by 100.  Find the percentile rank in the frequency histogram.


Q. For a person who scores 25 out of 30 on the conscientiousness scale, what is the percentile rank computed using the z-score? 

Click here (Mark-there is an error in this link) to sort the conscientiousness data and then compute the percentile rank from the sorted raw score distribution? 

What does a score of 25 out of 30 on the conscientiousness scale mean?

  the person is high on C.
  the person is about average on C.
  the person is low on C.
  the person is moderately high on C.


Q. If a person scores 20 out of 30 on the consciousness scale, what is the percentile rank computed using the z-score? using the raw score distribution? 

What does score of 20 out of 30 on the conscientiousness scale  mean?

  the person is high on C.
  the person is average on C.
  the person is moderately low on C.
  the person is moderately high on C.



Q. If a person scores 17 out of 30 on the conscientiousness scale, what is the percentile rank computed using the z-score? using the raw score distribution? 

What does score of 17 out of 30 on the conscientiousness scale  mean?

  the person is high on C.
  the person is average on C.
  the person is low on C.
  the person is very low on C.




 

Page 13. How is relative performance related to changes in the standard deviation and the sample mean?

Q. What if the standard deviation of the sample of conscientiousness scores doubled (ie., 6.34), but the mean stays constant at 24.1?  The larger standard deviation means that peoples C scores varied more from one another, and from the group mean.  Let's consider a score of 20 out of 30.  What is the percentile rank computed using the z-score? 

Q. What does score of 20 out of 30 on the conscientiousness scale mean in this transformed distribution?
 

  the person is high on C.
  the person is average to moderately low on C.
  the person is moderately high on C.
  the person is very low on C.
Q. Now, what if the standard deviation remains the same as the original sample (ie., 3.17), but the sample mean changes and is equal to 26. Lets reconsider what a score of 20 out of 30 means.

  the person is high on C.

  the person is average on C.

  the person is moderately low on C.

  the person is very low on C.


  Q. Now, what if the standard deviation remains the same as the original sample (i.e., 3.17) but the sample mean changes and is equal to 15. Lets reconsider what a score of 20 out of 30 means.   the person is high on C.
  the person is average on C.
  the person is moderately low on C.
  the person is very low on C.
So, you can see that the meaning of a score depends on two related things. First, it depends on the mean and standard deviation of the sample, and second, on the probability of finding the score in the sample. The same score can indicate a high or low level of conscientiousness depending on the sample characteristics.

Page 14.  Compute your own percentile rank.

Now, we are ready to assess whether your score on the conscientiousness scale is high or low. First, lets begin by looking at the sample distribution once again.  Click here link to charting applet with conscientiousness data open to access the sample distribtution of conscientiousness scores.  Compute your percentile rank using the raw score distribution.  Enter your percentile rank here:  create field to enter numbers

The distribution has a mean of 24.10 and a standard deviation of 3.17;  compute a z score for your raw score.  Then use your z score to compute probability using the standard normal distribution.

Your z score is? 

The probability of someone scoring lower than you on conscientiousness is? 

The probability of someone scoring higher than you on conscientiousness is? 


Sign in so your instructor knows that you have completed this module.

 
 
 

Eagle ID number: 

Course Number: