Deck 2: Statistics Review for PSYC Measurement

Question 1

Shaun completed his first exam in biology and earned a score of 38. This seemed low to him, so he asked his professor what percentile rank this represented. The professor replied it was equivalent to the 85th percentile. What is the interpretation of Shaun's score (a percentile rank score of 85) in this case?

A) He scored higher than 15% of the other students taking the test.
B) He scored higher than 85% of the other students taking this test.
C) He actually did even worse than he thought on his first biology exam!
D) Unfortunately, percentile rank scores tell us nothing informative about an individual's test score.

Answer 1

B

Question 2

Your study partner can't understand why Professor Shultz spent so much time discussing z scores, in that they don't seem all that important. What would be your response to your study partner as to why z scores are so important in the field of tests and measurement?

A) They allow for both positive and negative scores.
B) They are easily understood by non-testing professionals.
C) They allow comparisons across different distributions of scores.
D) They always carry a positive connotation regarding test performance.

Answer 2

C

Question 3

If I wanted to estimate a test takers true underlying level of ability that resulted from a given assessment, I would compute a(n) _________.

A) t-test
B) alpha reliability
C) correlation coefficient
D) 95% confidence interval

Answer 3

D

Question 4

A statistic is to a sample, as a(n) __________.

A) parameter is to a population
B) upper limit is to a lower limit
C) interval scale is to an ordinal scale
D) dependent variable is to an independent variable

Question 5

Greek letters are used in statistics to _________.

A) designate characteristics of a sample
B) designate characteristics of a population
C) describe a sample of scores larger than 100
D) imply that a normal distribution is being used in the study

Question 6

A frequency distribution of test scores with a mean of 100 and a median of 90 would be described as _______.

A) negatively skewed
B) positively skewed
C) symmetrical or normal
D) unable to determine unless you also know the standard deviation

Question 7

The preferred measure of central tendency for test scores that are measured with an interval level of measurement, which are not severely skewed, would be ______.

A) the mode
B) the median
C) arithmetic mean
D) harmonic mean

Question 8

Which of the following words could legitimately fit into this sentence: "The simple frequency distribution has two _____ of 13 and 18".

A) means
B) models
C) medians
D) all three alternatives could be correct

Question 9

Assume you conduct a study to evaluate the relationship between the amount of time a child is read to at age 3 and their reading ability at age 15. You find a correlation of .04. This correlation suggests _____.

A) unable to tell without knowing the number of participants in the study
B) a moderate relationship; as reading to children increases, reading ability decreases
C) a very strong relationship; as reading to children increases, reading ability increases
D) there is almost no relationship between reading to a child and their later reading ability

Question 10

How would you interpret a test score of z=+1.0? What percentile score would this represent?

Question 11

Why do Shultz, Whitney, and Zicker (2021) suggest in Module 2 that descriptive statistics tend to be more informative in applied psychological testing than statistical significance testing?

Question 12

Why do Shultz, Whitney, and Zicker (2021) suggest that in terms of inferential statistics in psychological testing, that estimation is more informative than statistical significance testing?

Question 13

What are the major determinants in deciding whether to construct a simple (i.e., individual score) frequency distribution or a grouped frequency distribution to describe a set of test scores?

Question 14

Why is it always important to first construct a scatter plot when examining bivariate relationships, before then computing the bivariate correlation coefficient?