Week 7

Question 1

Direction of Correlation

Answer 1

If r is above 0 (1>r>0), correlation is positive
If r is below 0, (-1<r<0), correlation is negative

Question 2

Strength of Correlation

Answer 2

If r is close to 1 (r - 1), correlation is strong
If r is close to 0 (r - 0), correlation is weak

Question 3

Variance of Correlation

Answer 3

If r2 is close to 1 (r2 - 1), correlation explains a lot of variance
If r2 is close to 0 (r2 - 0), correlation explains only a little variance
1-r2 is the amount of variance not explained (noise or error)

Question 4

Regression Equation

Answer 4

y = m*x + c
When x = 0, y = intercept
When x increases by 1, y increases by the slope

Question 5

Reporting Correlations

Answer 5

r((N-2)) = (Pearson’s r), p(p-value)
Eg (N = 66, Pearson’s r = 0.881, p-value = <0.001)
Therefore, : r(64)= .881, p<0.001

Question 6

Reporting Regressions

Answer 6

F((df1),(df2)) = (F-value), p=(p-value)
Eg (df1 = 1, df2 = 64, F-value = 233, p-value = <0.01)
Therefore, : F(1,64)=233, p<0.01)

Question 7

Correlation

Answer 7

Describes single strength & direction of Relationship
Linear Relationship
X & Y axis are inter-changeable
Does not allow prediction

Question 8

Regression

Answer 8

Describes multiple directions & strengths of relationships
Linear Relationship
X & Y axis are not inter-changeable
Allows for prediction

Question 9

Multiple Regression Predictors

Answer 9

Predictors can be nominal, ordinal or discrete
Normally-distributed or not
Linear or Non-linear

Question 10

Problems with Correlation & Regression

Answer 10

Correlation does not equal causation
Non-linear relationships can cause problems
Extrapolation (continuing trend) is wrong

Question 11

Solutions to problems in correlation & regression

Answer 11

Look at the data
Check for mistakes
Transform the Data
- Quadratic
- Cubic
Logarithmic