PSY 370 EXAM 2 QUESTIONS & VERIFIED ANSWERS
Why do test users calculate standard scores? What are they useful for? - Answers -
Transformations of raw scores are done so that they have a particular mean and SD
-its gives content and makes it easier to interpret a single score
What are linear transformations? - Answers - -when we transform raw test scores, we
create a more informative scale to help us interpret a particular score in terms of where
the score fits in an overall distribution of test scores
-Linear: change the unit of measurement, but do not change the characteristics of the
raw data in any way
What are area transformations? - Answers - Area: change not only the unit of
measurement, but also the unit of reference
-rely on the normal curve
-magnify the differences between individuals at the middle of the distribution and
compress differences between individuals at extremes of the distribution
What is the difference between linear and area transformations? - Answers - Area:
change not only the unit of measurement, but also the unit of reference
-Linear deals with percentages, standard deviations units, z scores and t scores
-Area deals with percentiles and stanine
What are Z-scores, and how do you calculate them? - Answers - -similar to a SD units
except that it is represented as a whole number with a decimal point
-helps us understand how many SD an individual test score is above or below the
distribution mean
-distribution shape stays the same
Mean= 0
SD=1
z score= x-μ /σ
x=raw score
μ = mean
σ= SD
Be able to interpret various z-scores. - Answers -
Why does it matter whether a z-score is positive or negative? - Answers - Yes.
Positive is above the mean and a negative score indicating it is below the mean.
,What are T-scores, and how do you calculate them? - Answers - similar to z scores;
help us understand how many SDs an individual test score is above or below the
distribution mean
-but are different because they always have mean of 50 and a SD of 10
-ALWAYS positive unlike z scores
-need to know z score
T score= (z times 10) +50
Be able to interpret various t-scores. - Answers -
****Why might you use t-scores instead of z-scores? - Answers - -you can transform
scores to fit any mean and SD you want this way
IQ: Mean 100, Sd=15
What are percentiles? - Answers - -allow us to determine a persons relative standing
compared with others in a distribution of scores
-percentage of scores in a distribution that fall at or below a given raw score (percentile
rank)
-percentage of scores in a distribution falls below a given raw score
How do you calculate them? - Answers - (B + .5E)/n times 100
B= number of scores below the individuals score
E= number of scores equal to the individuals score
n=the number of people who took the test
1. sort the distribution of scores from lowest to highest
2. count the number of scores that fall below the raw score of interest (B)
3. Count the number of scores equal to the raw score of interest (E)
4. count the total number of scores (n)
5. Substitute the values obtained in steps 2,3,4 into formula for percentile rank
Be able to interpret various percentile scores. - Answers -
What are stanines? - Answers - -are a standard score scale with nine points that allows
us to describe a distribution in words instead of numbers (from 1= very poor to 9=very
superior)
-expressed in whole numbers 1-9
ex: 1,2,3-represent performance below mean
ex: 4,5,6-considered to be average or close to the mean
7, 8-above average
9-thought as exceptional
How do you find them (very generally speaking)? - Answers - 9 buckets-the bucket a
person is in is their stannic value (1-9)
, Why might you use them instead of another standard score format? - Answers -
differenciates more between people in the middle than z scores or SD units
What is the difference between the normative approach to scoring and the criterion
approach? - Answers - Normative: Transformed scores give us relative information -
how does the test-taker compare to the group?
-Norm-referenced scoring.
Criterion: the measure of performance that we correlate with test scores
-Sometimes we are more interested in absolute information - how does the test-taker
compare to a fixed standard?
-Criterion-referenced scoring.
-For correct answers, percentage scores are common.
-Cut scores
-Can be set with reference to some empirical criterion...
-or in other ways, such as by expert judgment.
***Give and recognize examples of each approach (normative and criterion): - Answers
- Example of criterion: if higher test scores relate to higher performance ratings, the test
shows evidence of validity based on the relationship between these two variables
(criterion related validity)
-GRE scores to predict graduate school success
Example of normative: The normative approach compares test-takers to each other on
a specific test while the criterion approach compares your measure of interest to the
scores of other scales to see if your test behaves as you'd expect it to with other
constructs. An example of a normative approach would be to compare an individual's
scores on an IQ test to the population norms to see how they compare to "average."
When would you use normative approach? - Answers -
When would you use criterion approach? - Answers - for predictive and concurrent
validity
What are norms? - Answers - norms are the average scores of some predefined group
we want to compare to
-Norm referenced tests are given to a clearly defined norm group
e.g., Colorado third graders; clearly defined group
-Stats for the norm group are then included in the test manual
-often includes conversion to percentile ranks
-other options include age or grade norms
-possibly race or cultural norms
-need to consider the composition of the norm group to know how to interpret test score
Why do test users calculate standard scores? What are they useful for? - Answers -
Transformations of raw scores are done so that they have a particular mean and SD
-its gives content and makes it easier to interpret a single score
What are linear transformations? - Answers - -when we transform raw test scores, we
create a more informative scale to help us interpret a particular score in terms of where
the score fits in an overall distribution of test scores
-Linear: change the unit of measurement, but do not change the characteristics of the
raw data in any way
What are area transformations? - Answers - Area: change not only the unit of
measurement, but also the unit of reference
-rely on the normal curve
-magnify the differences between individuals at the middle of the distribution and
compress differences between individuals at extremes of the distribution
What is the difference between linear and area transformations? - Answers - Area:
change not only the unit of measurement, but also the unit of reference
-Linear deals with percentages, standard deviations units, z scores and t scores
-Area deals with percentiles and stanine
What are Z-scores, and how do you calculate them? - Answers - -similar to a SD units
except that it is represented as a whole number with a decimal point
-helps us understand how many SD an individual test score is above or below the
distribution mean
-distribution shape stays the same
Mean= 0
SD=1
z score= x-μ /σ
x=raw score
μ = mean
σ= SD
Be able to interpret various z-scores. - Answers -
Why does it matter whether a z-score is positive or negative? - Answers - Yes.
Positive is above the mean and a negative score indicating it is below the mean.
,What are T-scores, and how do you calculate them? - Answers - similar to z scores;
help us understand how many SDs an individual test score is above or below the
distribution mean
-but are different because they always have mean of 50 and a SD of 10
-ALWAYS positive unlike z scores
-need to know z score
T score= (z times 10) +50
Be able to interpret various t-scores. - Answers -
****Why might you use t-scores instead of z-scores? - Answers - -you can transform
scores to fit any mean and SD you want this way
IQ: Mean 100, Sd=15
What are percentiles? - Answers - -allow us to determine a persons relative standing
compared with others in a distribution of scores
-percentage of scores in a distribution that fall at or below a given raw score (percentile
rank)
-percentage of scores in a distribution falls below a given raw score
How do you calculate them? - Answers - (B + .5E)/n times 100
B= number of scores below the individuals score
E= number of scores equal to the individuals score
n=the number of people who took the test
1. sort the distribution of scores from lowest to highest
2. count the number of scores that fall below the raw score of interest (B)
3. Count the number of scores equal to the raw score of interest (E)
4. count the total number of scores (n)
5. Substitute the values obtained in steps 2,3,4 into formula for percentile rank
Be able to interpret various percentile scores. - Answers -
What are stanines? - Answers - -are a standard score scale with nine points that allows
us to describe a distribution in words instead of numbers (from 1= very poor to 9=very
superior)
-expressed in whole numbers 1-9
ex: 1,2,3-represent performance below mean
ex: 4,5,6-considered to be average or close to the mean
7, 8-above average
9-thought as exceptional
How do you find them (very generally speaking)? - Answers - 9 buckets-the bucket a
person is in is their stannic value (1-9)
, Why might you use them instead of another standard score format? - Answers -
differenciates more between people in the middle than z scores or SD units
What is the difference between the normative approach to scoring and the criterion
approach? - Answers - Normative: Transformed scores give us relative information -
how does the test-taker compare to the group?
-Norm-referenced scoring.
Criterion: the measure of performance that we correlate with test scores
-Sometimes we are more interested in absolute information - how does the test-taker
compare to a fixed standard?
-Criterion-referenced scoring.
-For correct answers, percentage scores are common.
-Cut scores
-Can be set with reference to some empirical criterion...
-or in other ways, such as by expert judgment.
***Give and recognize examples of each approach (normative and criterion): - Answers
- Example of criterion: if higher test scores relate to higher performance ratings, the test
shows evidence of validity based on the relationship between these two variables
(criterion related validity)
-GRE scores to predict graduate school success
Example of normative: The normative approach compares test-takers to each other on
a specific test while the criterion approach compares your measure of interest to the
scores of other scales to see if your test behaves as you'd expect it to with other
constructs. An example of a normative approach would be to compare an individual's
scores on an IQ test to the population norms to see how they compare to "average."
When would you use normative approach? - Answers -
When would you use criterion approach? - Answers - for predictive and concurrent
validity
What are norms? - Answers - norms are the average scores of some predefined group
we want to compare to
-Norm referenced tests are given to a clearly defined norm group
e.g., Colorado third graders; clearly defined group
-Stats for the norm group are then included in the test manual
-often includes conversion to percentile ranks
-other options include age or grade norms
-possibly race or cultural norms
-need to consider the composition of the norm group to know how to interpret test score
