Correlations Flashcards
Statement 1: The significance test you get with a correlation coefficient tells you whether it is significantly DIFFERENT from zero.
Statement 2: The greater the degree of scatter in a scatterplot between two variables (both with Gaussian distributions), the larger the correlation coefficient between those two variables will be.
(a) Both statements are true.
(b) Statement 1 true; Statement 2 false.
(c) Statement 1 false; Statement 2 true.
(d) Both statements are false.
The answer was b. See the online correlation lecture. Statement 1 is true but Statement 2 is false (because the greater the scatter, the SMALLER the correlation coefficient will be).
Statement 1: If a validity coefficient is large enough to be of practical importance, then it is not necessary for it to be statistically significant.
Statement 2: Spearman’s Rho is a type of correlation coefficient appropriate for use on nominal data.
(a) Both statements are true.
(b) Statement 1 true; Statement 2 false.
(c) Statement 1 false; Statement 2 true.
(d) Both statements are false.
The answer was d. See the online correlation lecture. Statement 1 is false. If a correlation coefficient from a sample is not statistically significant then we can’t be sure the population correlation isn’t zero, no matter how big the coefficient is (this might happen if there were, for example, three people in your sample). Statement 2 is false. Spearman’s Rho is suitable for ordinal data not nominal data: you can’t do a correlation on nominal data (the very concept is meaningless). Instead you’d do a chi square test or similar.
Statement 1: The confidence interval associated with a correlation coefficient decreases as the sample size increases.
Statement 2: If we’re expecting a certain population correlation to be large then there would never be any point in sampling more than a small number of people when we attempt to estimate it
(a) Both statements are true.
(b) Statement 1 true; Statement 2 false.
(c) Statement 1 false; Statement 2 true.
(d) Both statements are false.
The answer was b. See video lecture on correlations. Statement 1 is true. The confidence interval decreases as the sample size increases because the estimate becomes more accurate the more people you sample. This means the confidence interval is smaller (i.e. there’s a smaller margin of error in our attempt to estimate of the population correlation) with larger samples. Statement 2 is false. Even if we were expecting a large correlation coefficient, a larger sample will allow us to estimate the magnitude of this correlation more accurately (even though we might not need the larger sample to deternine whether the correlation was significantly greater than zero).
The correlation coefficient tells you:
(a) The strength of the relationship
(b) The main effect
(c) The direction of the relationship
(d) Both a and c
D. Both a and c.
Statement 1: If your data are appropriate for a parametric test, variables must be normally distributed and be on an ordinal scale of measurement.
Statement 2: Spearman’s Rho uses rank order of the data rather than raw scores and therefore variables don’t need to be normally distributed
(a) Both statements are true.
(b) Statement 1 true; Statement 2 false.
(c) Statement 1 false; Statement 2 true.
(d) Both statements are false.
Correct answer: C. See Correlation Lecture. Slide 36. Statement 1 is false because variables must be on an interval scale of measurement in order to use a parametric test.
Statement 1: The strength of a correlation would be affected when any linear transformation is done on any of the measures.
Statement 2: The correlation output from SPSS is an exact value of the degree to which two variables are correlated in the real world.
(a) Both statements are true
(b) Statement 1 is true; Statement 2 is false
(c) Statement 1 is false; Statement 2 is true
(d) Both statements are false
Correct answer: D.
Explanation: see online correlations lecture, slides 20 (Statement 1) & 21 (Statement 2).
Statement 1: Strength of a correlation wouldn’t be affected because they are all standardised (not raw numbers). Statement 2: The correlation output from SPSS is only an estimate based upon the sample we had, it is not the exact value inherent in the population.
Statement 1: The magnitude of the correlation coefficient is not dependent on the amount of noise in the relationship.
Statement 2: Spearman’s rho is a more sensitive non-parametric test for a correlation than Pearson’s r.
a) Both statements are true
b) Statement one is true, statement two is false
c) Statement one is false, statement two is true
d) Both statements are false
Correct answer: D.
Explanation: See Correlation Lecture, slides 12 (Statement 1) & 36 (Statement 2). Statement 1: The correlation magnitude is dependent on noise. Statement 2: Spearman’s rho is a non-parametric test, however it is less sensitive than Pearson’s r.
What does NOT influence the strength of a correlation?
(a) A sample size
(b) The noise/scatter
(c) Measurement scale
(d) The linearity of the relationship
Correct answer: C
A correlation is a standardised measure of the relationship between two variables and is therefore independent of the actual scales they were measured upon.
If we can predict someone’s job performance with 100% accuracy from their score on a test, what does this indicate?
a) A perfect positive correlation
b) The test is normally distributed
c) A perfect correlation
d) It has excellent reliability
Correct answer: C.
Explanation: See Correlation Lecture, slide 8. It is a perfect correlation as from this statement we cannot deduce the direction, therefore it could be a positive or negative correlation.
Statement 1: A correlation can be calculated if there is zero variability.
Statement 2: If both novices and experts were given an easy task in a study this may restrict the range of scores.
a) Both statements are true
b) Statement 1 is true; Statement 2 is false
c) Statement 1 is false; Statement 2 is true
d) Both statements are false
Correct answer: C.
Explanation: See Correlation Lecture, slides 14 (Statement 1) & 15 (Statement 2). Statement 1: You need variation to have a correlation. Statement 2: All participants would likely score highly on the task, therefore there would be minimal variability, resulting in restriction of range.
Statement 1: To obtain a population correlation, a small sample should be tested to yield a small margin of error.
Statement 2: A restriction of range on your data will change the size of the correlation coefficient.
a) Both statements are true
b) Statement 1 is true and statement 2 is false
c) Statement 1 is false and statement 2 is true
d) Both statements are false
Correct answer: C.
Explanation: See correlations lecture, slide 24 (Statement 1) & slide 15 (Statement 2).
Statement 1 is false because a population correlation obtained by a small sample will have a large margin of error. Statement 2 is true because when we restrict the range of the data, we only take into account a small part of the population variance that will suppress the correlation.
Statement 1: A larger sample size will cause a stronger statistically significant correlation
Statement 2: To determine how many people are required to detect a particular correlation size, you can use a technique called parametric analysis. Both statements true Statement 1 true, Statement 2 false Statement 1 false, Statement 2 true Both statements false
Both Statements are False.
Statement 1: A large sample size can result in a weak statistically significant correlation; size does not always determine the strength. Conversely, you can have a strong correlation that is not statistically significant in small sample sizes. [online correlation lecture, slide 32]
Statement 2: The technique is a “Power” analysis, not a parametric analysis. A parametric is related to the distribution of frequencies within an analysis. [online correlation lecture,slide 33]
Statement 1: A correlation between two measures can vary between -1 and 1.
Statement 2: The significance test you get with a correlation coefficient tells you whether it differs significantly from zero.
(a) Both statements are true
(b) Statement 1 is true; Statement 2 is false
(c) Statement 1 is false; Statement 2 is true
(d) Both statements are false
Correct answer: A
Explanation: See online correlation lecture, slide 3 (Statement 1) & 21 (Statement 2).
Statement 1 is true as correlations vary between -1 and +1 and the number is known as the correlation coefficient. Statement 2 is true as the coefficient only tells if it differs from zero.
Power analysis is:
a) A test to detect a particular size of correlation as significant
b) A test to detect how many people you need to test in order to detect a particular size of correlation as significant
c) A test to detect the likelihood of a significant result when the null hypothesis is true
d) A test to determine the effect size of a correlation
Correct answer: B.
Explanation: See Correlation Lecture, slide 33. Power analysis is used to determine how many people are needed for a test.
A correlation coefficient is:
(a) A descriptive statistic
(b) A measure of effect size
(c) Both a descriptive statistic and a measure of effect size
(d) Neither a descriptive statistic and a measure of effect size
Correct answer: C
Explanation: See online correlation lecture, slide 3.
It is a descriptive statistic because it describes the strengths and direction of the relationship. It’s a measure of effect size because it gives an indication of the strength of relationship (i.e., effect size).
