Chapter 13: Regression and Correlation Flashcards
What is correlation and regression?
Statistical techniques are often used in relation with one another
What can you do with correlation and regression?
quantify the strength and direction of an association between variables
What is covary?
Two random variables are associated
What is regression?
For any given value of x we can predict the corresponding value of y, specific predictions based on an association
What is the correlation coefficient?
Number that quantifies the strength and direction of an association between variables
What numbers are correlations between?
1 and -1
What are the measures of association?
The variety of correlation coefficients
What do correlations measure?
Relationship between 2 variables (x and y)
Suppose that you have access to hospital records on patient age and their number of days spent in hospital. Your most appropriate measure of association would likely be ________.
Pearsons r
What can you say if no pattern appears to exist between variables?
Statistically independent
What is the strength variation in correlation?
0 and +- 1
What does a stronger pattern or association mean?
Larger the correlation coefficient
There is no such thing as a zero correlation.
False
A correlation coefficient distinguishes between independent and dependent variables.
false
Are a coefficient of -.35 and +.35 equally strong?
Yes can think of it as a continuum
What is the relationship between positive and negative correlations?
Positive go together, in negative they move apart
What do scatterplots plot?
Interval or ratio level data onto a graph in a case-by-case manner
What does the horizontal axis represent?
Predictor variable
What does the vertical axis stand for?
Criterion variable
What is the line of best fit?
straight line that best represents data on scatter plot (regression line), nature of relation between two variables, eyeball method, decides if it is linear
What happens if a straight line can’t be drawn through the data points?
There may be no relationship at all
What are scatterplots very useful for spotting?
Outliers in data
What do outliers have a significant impact on?
Correlation coefficients
What do scatterplots tell us in regard to calculations?
Whether or not calculating pearsons r is appropriate
What are the requirements for pearsons r?
Random sampling, continuous interval or ratio data, normally distributed variables, no significant outliers, linear
What can pearsons r not detect?
Non-linear associations
Study The Who grows up to be a philanderer thing cause it is mentioned a LOT
What is a advantage of Pearson coefficients?
Possess a common metric, compare strength of relationships with one another
What is the remaining percentage 100-r^2?
Unexplained variation
What is unexplained variation?
Variation in y is accounted for by some variable that isn’t x
What do you use when you cant use Pearsons r?
Rank correlation
What is monotonicity?
Whether or not one set of scores tends to increase or decrease along with another set (linear)
What are non-monotonic associations?
Scores can’t increase together then decrease
What is a spurious correlation?
Association is caused by a third variable that affects both the associations two variables simultaneously;
What is reverse causation?
Might think x causes y but actually y causes x
What is the post-hoc fallacy?
If B followed A then A must have caused B
What happens with stronger associations?
The more likely it is causal
What is empirical support?
Gathered evidence using our senses
What is the slope?
Steepness of the line of best fit, amount of change in y for every change in x (descriptive statistic)
What is the intercept?
Point regression line crosses y axis when x=0
What are the steps of the regression equation?
Calculate slope (b), calculate intercept (a), using value of x solve for y