Chapter 5: z-Scores: Location of Scores and Standardized Distributions

Chapter 5: z-Scores: Location of Scores and Standardized Distributions I t d ti Introduction to t z-Scores S • • • • In the previous two chapters, we introduced th concepts the t off the th mean and d the th standard t d d deviation as methods for describing an entire distribution of scores. Now we will shift attention to the individual scores within a distribution. In this chapter, we introduce a statistical technique that uses the mean and the standard deviation to transform each score (X value) into a z-score or a standard score. The purpose of z-scores, or standard scores, is to identify and describe the exact location of every score in a distribution. distribution I t d ti Introduction to t z-Scores S cont. t • In other words, the process of transforming X values into z-scores serves two useful purposes: – Each z-score tells the exact location of the original X value within the distribution. – The z-scores form a standardized distribution that can be directly y compared to other distributions that also have been transformed into z-scores. ZS Z-Scores and dL Location ti in i a Distribution Di t ib ti • • • One of the primary purposes of a z-score is to describe the exact location of a score within a distribution. The z-score accomplishes this goal by transforming each X value into a signed number (+ or -) so that: – The sign tells whether the score is located above ((+)) or below (-) ( ) the mean, and – The number tells the distance between the score and the mean in terms of the h number b off standard d dd deviations. i i Thus, in a distribution of IQ scores with μ= 100 and σ = 15, a score of X = 130 would be transformed into z = +2.00. +2 00 Z-Scores and Location in a Distribution cont. • • • The z value indicates that the score is located above the mean (+) by a distance of 2 standard deviations (30 points). Definition: A z-score specifies the precise location of each X value within a distribution. – The sign of the z-score (+ or -) signifies whether the score is above the mean (positive) or below the mean (negative). – The numerical value of the z-score specifies the distance from the mean b counting by i the h number b off standard d d deviations between X and μ. Notice that a z-score always consists of two parts: a sign (+ or -)) and a magnitude. magnitude Z-Scores and Location in a Distribution cont. • • • • • Both parts are necessary to describe completely where a raw score is located within a distribution. Figure 5.3 shows a population distribution with various positions identified by their z-score values. Notice that all z-scores above the mean are p positive and all z-scores below the mean are negative. The sign of a z-score tells you immediately whether the score is located above or b l below the h mean. Also, note that a z-score of z =+1.00 corresponds to a position exactly 1 standard deviation above the mean. mean Z-Scores and Location in a Distribution cont. • • • • A z-score of z = +2.00 is always located exactly 2 standard deviations above the mean. The numerical value of the z-score tells you the number of standard deviations from the mean. Finally, you should notice that Figure 5.3 does not g give any y specific p values for the population mean or the standard deviation. The locations identified by z-scores are the h same for f all ll distributions, di ib i no matter what mean or standard deviation the distributions may have. Fig. 5-3, p. 141 z-Score S F Formula l • The formula for transforming scores into z-scores is Formula 5.1 • • The numerator of the equation, X - μ, is a d i ti score (Chapter deviation (Ch t 4, 4 page 110); 110) it measures the distance in points between X and μ and indicates whether X is located above or below the mean. The deviation score is then divided by σ because we want the z-score to measure distance in terms of standard deviation units. i Determining g a Raw Score ((X)) from a z-Score • • Although the z-score equation (Formula 5.1…slide #9) works well for transforming X values into z-scores, it can be awkward when you are trying to work in the opposite direction and change z-scores back into X values. The formula to convert a z-score into a raw score (X) is as follows: Formula 5.2 Determining g a Raw Score ((X)) from a z-Score cont. • • In the formula (from the previous slide), the value of zσ is the deviation of X and determines both the direction and the size of the distance from the mean. Finally you should realize that Formula Finally, 5.1 and Formula 5.2 are actually two different versions of the same equation. Other Relationships p Between z,, X,, μ, and, σ • • • • In most cases, we simply transform scores (X values) into z-scores, or change z-scores back into X values. However, you should realize that a z-score establishes a relationship between the score, the mean, and the standard deviation. This relationship p can be used to answer a variety of different questions about scores and the distributions in which they are located. Pl Please review i the h following f ll i example l (next ( slide). Other Relationships p Between z,, X,, μ, and, σ cont. • • In a population with a mean of μ = 65, a score of X = 59 corresponds to z = -2.00. What is the standard deviation for the population? – To T answer the th question, ti we begin b i with ith the z-score value. – A z-score of -2.00 indicates that the corresponding score is located below the mean by a distance of 2 standard deviations. – By simple subtraction, you can also determine that the score (X = 59) is located below the mean (μ = 65) by a distance of 6 points. Other Relationships p Between z,, X,, μ, and, σ cont. • Thus, 2 standard deviations correspond to a distance of 6 points, which means that 1 standard deviation must be σ = 3. Using g z-Scores to Standardize a Distribution • • • • It is possible to transform every X value in a distribution into a corresponding z-score. The result of this process is that the entire distribution of X values is transformed into a distribution of z-scores z scores (Figure 5.5). 5 5) The new distribution of z-scores has characteristics that make the z-score transformation a very y useful tool. Specifically, if every X value is transformed into a z-score, then the distribution of z-scores will have the following properties: Using g z-Scores to Standardize a Distribution cont. • Shape – The shape of the z-score distribution will be the same as the original distribution of raw scores. – If the th original i i l di distribution t ib ti is i negatively ti l skewed, for example, then the z-score distribution will also be negatively skewed. – In other words, transforming raw scores into z-scores does not change anyone's position in the distribution. – Transforming a distribution from X values to z values does not move scores from one position to another; the procedure simply relabels each score (see Figure 5.5). Using g z-Scores to Standardize a Distribution cont. • The Mean – The Th z-score distribution di ib i will ill always l have a mean of zero. – In Figure 5.5, the original distribution of X values has a mean of μ = 100. 100 – When this value, X=100, is transformed into a z-score, the result is: – Thus,, the original g population p p mean is transformed into a value of zero in the z-score distribution. – The fact that the z-score distribution h a mean off zero makes has k it easy to t identify locations Fig. 5-5, p. 146 Using g z-Scores to Standardize a Distribution cont. • The Standard Deviation – The distribution of z-scores will always have a standard deviation of 1. – Figure 5.6 demonstrates this concept with a single i l distribution di t ib ti that th t has h two t sets t off labels: the X values along one line and the corresponding z-scores along another line. – Notice that the mean for the distribution of z-scores is zero and the standard deviation is 1. – When any distribution (with any mean or standard deviation) is transformed into z-scores, the resulting distribution will always have a mean of μ = 0 and a standard deviation of σ = 1. Using g z-Scores to Standardize a Distribution cont. • • Because all z-score distributions have the same mean and the same standard deviation, the z-score distribution is called a standardized distribution. Definition: A standardized distribution is composed of scores that have been transformed to create predetermined values for μ, and, σ. • Standardized distributions are used to make dissimilar distributions comparable. Using g z-Scores for Making g Comparisons • • One advantage of standardizing distributions is that it makes it possible to compare different scores or different individuals even though they come from completely different distributions. Normally, if two scores come from different distributions, it is impossible to make any direct comparison between them. – Suppose, for example, Bob received a score of X=60 on a psychology exam and a score of X=56 on a biology test. test – For which course should Bob expect the better grade? Using g z-Scores for Making g Comparisons cont. – Because the scores come from two different distributions, you cannot make any direct comparison. – Without additional information, it is even impossible to determine whether Bob is above or below the mean in either distribution. – Before y you can begin g to make comparisons, you must know the values for the mean and standard deviation for each distribution. – Instead I d off drawing d i the h two distributions to determine where Bob's two scores are located, we p y can compute p the two z-scores simply to find the two locations. Using g z-Scores for Making g Comparisons cont. – For psychology, Bob's z-score is: – For biology, Bob's z-score is: – Note that Bob's z-score for biology is +2.0, which means that his test score i 2 standard is d dd deviations i i above b the h class mean. – On the other hand, his z-score is +1.0 for psychology, psychology or 1 standard deviation above the mean. Using g z-Scores for Making g Comparisons cont. – In terms of relative class standing, Bob is doing much better in the biology class. – Notice that we cannot compare Bob's two exam scores (X = 60 and X = 56) because the scores come from different distributions with different means and standard deviations. – However, we can compare the two z-scores because all distributions of z-scores have the same mean (μ = 0) and the same standard deviation (σ = 1). 1) C Computing ti z-Scores S f Samples for S l • • • Although z-scores are most commonly used in the context of a population, the same principles can be used to identify individual locations within a sample. The definition of a z-score z score is the same for a sample as for a population, provided that you use the sample mean and the sample standard deviation to specify each z-score location. Thus, for a sample, each X value is transformed into a z-score so that – The Th sign i off the h z-score indicates i di whether the X value is above (+) or below (-) the sample mean, and C Computing ti z-Scores S f Samples for S l cont. t – The numerical value of the z-score identifies the distance between the score and the sample mean in terms of the sample standard deviation. – Expressed as a formula formula, each X value in a sample can be transformed into a z-score as follows: • Similarly, y, each z-score can be transformed back into an X value, as follows: St d di i Standardizing a Sample S l Distribution Di t ib ti • • • If all the scores in a sample are transformed into z-scores, the result is a sample of z-scores. The transformation will have the same properties that exist when a population of X value is transformed into z-scores. Recall this is true for data collected from ap population, p , as well ((Refer to slide #16 for the same on population distributions). – Specifically, • The sample of z-scores will have the same shape as the original sample of scores. • The sample of z-scores will have a mean of M=0. M 0 Standardizing g a Sample p Distribution cont. • The sample of z-scores will have a standard deviation of s = 1. – Note that the set of z-scores is still considered to be a sample (Just like the set of X values) and the sample formulas must be used to compute variance and standard deviation.