Standardization allows the researcher to:

Question

Standardization allows the researcher to: (select all that apply)

A. Identify errors in the data

B. Normalize the data

C. Compare the score to others scores in the distribution

D. Compare scores across different scales

 

I cant find this in my text and do not remember it being discussed in class.

Answer 1

Casey:

 

When you compute a standardized score for a given value, you subtract the mean from the value given and divide the difference by the standard deviation.  The sign of the difference is retained, so that you can keep track of whether the value was below the mean or above it.  The standardized score is a number WITHOUT any units, just a pure unitless number.

 

It is used for comparison of values from different situations and different scales.  For example, suppose that  Barbra Swift has a score of 69 in a test in music that had a mean of 59 and SD = 5, while Taylor Streisand has a score of 75 in a Math test that had a mean of 70 and SD 3.  Whose performance was better? The scores 59 and 75 are not directly comparable, because one is in Music and the other is in Math and the students taking the tests may be quite different.  So their standardized scores are calculated and compared.  Barbra's is (69-59)/5 = 2.00 and Taylor's is (75-70)/3 = 1.67.  So relatively speaking, Barbra's performance was better.

 

As another example, suppose a car in Zurich, (Europe) gives a kilometers per liter average of 10 where the mean of such cars is 12.5 kpl with an SD of 2.  A car in the US gives an average of 25 mpg, where the mean of such cars is 28 mpg with a SD of 3.  Which car is under-performing?  The Zurich car's standardized score is -1.25, while the US car's standardized score is -1. So the Zurich car is relatively more underperforming than the US car.  Note that the standardized scores allowed this comparison when the SCALES were DIFFERENT.

 

My choice among the given options A, B, C, D is D

 

 

See that you do not need standardization if you want to compare values in the same distribution (Option C).  Post a comment if you have any questions or such.

 

Dr. G.