You are on page 1of 51

Correlation and

Regression
Correlation and Regression
The test you choose depends on level of measurement:

Independent Dependent Test

Dichotomous Interval-Ratio Independent Samples t-test
Dichotomous

Nominal Interval-Ratio ANOVA
Dichotomous Dichotomous

Nominal Nominal Cross Tabs
Dichotomous Dichotomous

Interval-Ratio Interval-Ratio Bivariate Regression/Correlation
Dichotomous

Correlation and Regression
Bivariate regression is a technique that fits a
straight line as close as possible between all the
coordinates of two continuous variables plotted
on a two-dimensional graph--to summarize the
relationship between the variables

Correlation is a statistic that assesses the
strength and direction of association of two
continuous variables . . . It is created through a
technique called regression

Bivariate Regression
For example:
A criminologist may be interested in the
relationship between Income and Number of
Children in a family or self-esteem and
criminal behavior.
Independent Variables
Family Income

Self-esteem
Dependent Variables
Number of Children

Criminal Behavior
Bivariate Regression
For example:
Research Hypotheses:
As family income increases, the number of children in
families declines (negative relationship).
As self-esteem increases, reports of criminal behavior
increase (positive relationship).
Independent Variables
Family Income

Self-esteem
Dependent Variables
Number of Children

Criminal Behavior
Bivariate Regression
For example:
Null Hypotheses:
There is no relationship between family income and the
number of children in families. The relationship statistic b = 0.
There is no relationship between self-esteem and criminal
behavior. The relationship statistic b = 0.
Independent Variables
Family Income

Self-esteem
Dependent Variables
Number of Children

Criminal Behavior
Bivariate Regression
Lets look at the relationship between self-
esteem and criminal behavior.
Regression starts with plots of coordinates of
variables in a hypothesis (although you will
hardly ever plot your data in reality).

The data:
Each respondent has filled out a self-esteem
assessment and reported number of crimes
committed.

Bivariate Regression
Y,
crime
s
X,
self-esteem
10 15 20 25 30 35 40
What do you think
the relationship is?
0



1



2




3




4




5




6




7




8




9



1
0


Bivariate Regression
Y,
crime
s
X,
self-esteem
10 15 20 25 30 35 40
Is it positive?
Negative?
No change?
0



1



2




3




4




5




6




7




8




9



1
0


Bivariate Regression
Y,
crime
s
X,
self-esteem
10 15 20 25 30 35 40
Regression is a procedure that
fits a line to the data. The slope
of that line acts as a model for
the relationship between the
plotted variables.
0



1



2




3




4




5




6




7




8




9



1
0


Bivariate Regression
Y,
crime
s
X,
self-esteem
10 15 20 25 30 35 40
The slope of a line is the change in the corresponding
Y value for each unit increase in X (rise over run).
Slope = 0, No relationship!
Slope = 0.2, Positive Relationship!
1
Slope = -0.2, Negative Relationship!
1
0.5
0



1



2




3




4




5




6




7




8




9



1
0


Bivariate Regression
The mathematical equation for a line:
Y = mx + b
Where: Y = the lines position on the
vertical axis at any point
X = the lines position on the
horizontal axis at any point
m = the slope of the line
b = the intercept with the Y axis,
where X equals zero
Bivariate Regression
The statistics equation for a line:
Y = a + bx
Where: Y = the lines position on the
vertical axis at any point (value of
dependent variable)
X = the lines position on the
horizontal axis at any point (value of
the independent variable)
b = the slope of the line (called the coefficient)
a = the intercept with the Y axis,
where X equals zero
^
^
Bivariate Regression
The next question:
How do we draw the line???

Our goal for the line:
Fit the line as close as possible to all the
data points for all values of X.
Bivariate Regression
Y,
crime
s
X,
self-esteem
10 15 20 25 30 35 40
How do we minimize the
distance between a line and all
the data points?
0



1



2




3




4




5




6




7




8




9



1
0


Bivariate Regression
How do we minimize the distance between a line and
all the data points?
You already know of a statistic that minimizes the
distance between itself and all data values for a
variable--the mean!
The mean minimizes the sum of squared
deviations--it is where deviations sum to zero and
where the squared deviations are at their lowest
value. (Y - Y-bar)
2

Bivariate Regression
The mean minimizes the sum of squared
deviations--it is where deviations sum to zero and
where the squared deviations are at their lowest
value.
Take this principle and fit the line to the place
where squared deviations (on Y) from the line are
at their lowest value (across all Xs).
(Y - Y)
2
Y = line
^ ^
Bivariate Regression
There are several lines that you could draw where the
deviations would sum to zero...
Minimizing the sum of squared errors gives you the
unique, best fitting line for all the data points. It is the
line that is closest to all points.
Y or Y-hat = Y value for line at any X
Y = case value on variable Y
Y - Y = residual
(Y Y) = 0; therefore, we use (Y - Y)
2
and minimize that!
^
^
^ ^
Bivariate Regression
Y,
crime
s
X,
self-esteem
10 15 20 25 30 35 40
0



1



2




3




4




5




6




7




8




9



1
0


Illustration of Y Y
= Yi, actual Y value corresponding w/ actual X
= Yi, line level on Y corresponding w/ actual X


5
-4
Y = 10, Y = 5

Y = 0, Y = 4


Bivariate Regression
Y,
crime
s
X,
self-esteem
10 15 20 25 30 35 40
0



1



2




3




4




5




6




7




8




9



1
0


Illustration of (Y Y)
2

= Yi, actual Y value corresponding w/ actual X
= Yi, line level on Y corresponding w/ actual X


5
-4
(Yi Y)
2
= deviation
2

Y = 10, Y = 5 . . . 25

Y = 0, Y = 4 . . . 16



Bivariate Regression
Y,
crime
s
X,
self-esteem
10 15 20 25 30 35 40
0



1



2




3




4




5




6




7




8




9



1
0


Illustration of (Y Y)
2

= Yi, actual Y value corresponding w/ actual X
= Yi, line level on Y corresponding w/ actual X
The goal: Find the line that minimizes
sum of deviations squared.


?
The best line will have the lowest value of sum of deviations squared
(adding squared deviations for each case in the sample.
Bivariate Regression
Y,
crime
s
X,
self-esteem
10 15 20 25 30 35 40
The fitted line for our
example has the equation:
Y = - X
, where e = distance from
line to data points or error
If you were to draw any other line, it
would not minimize . . .
(Y - Y)
2
0



1



2




3




4




5




6




7




8




9



1
0



Y = a + bX


e
Bivariate Regression
We use (Y - Y)
2
and minimize that!
There is a simple, elegant formula for
discovering the line that minimizes the sum of
squared errors
((X - X)(Y - Y))
b = (X - X)
2
a = Y - bX Y = a + bX
This is the method of least squares, it gives our
least squares estimate and indicates why we call
this technique ordinary least squares or OLS
regression
^
^
Bivariate Regression
Y
X
0 1
1





2





3





4





5





6





7





8





9




1
0


Considering that a regression line minimizes (Y - Y)
2
,
where would the regression line cross for an interval-ratio
variable regressed on a dichotomous independent variable?
^
For example:
0=Men: Mean = 6
1=Women: Mean = 4
Bivariate Regression
Y
X
0 1
1





2





3





4





5





6





7





8





9




1
0


The difference of means will be the slope.
This is the same number that is tested for
significance in an independent samples t-test.
^
0=Men: Mean = 6
1=Women: Mean = 4
Slope = -2 ; Y = 6 2X
Correlation
This lecture has
covered how to model
the relationship
between two
variables with
regression.
Another concept is
strength of
association.
Correlation provides
that.
Correlation
Y,
crime
s
X,
self-esteem
10 15 20 25 30 35 40
So our equation is:
Y = 6 - .2X
The slope tells us direction of
association How strong is
that?
0



1



2




3




4




5




6




7




8




9



1
0

^
Correlation
1





2





3





4





5





6





7





8





9




1
0


Example of Low Negative Correlation
When there is a lot of difference on the dependent variable across
subjects at particular values of X, there is NOT as much association
(weaker).
Y
X
Correlation
1





2





3





4





5





6





7





8





9




1
0


Example of High Negative Correlation
When there is little difference on the dependent variable across
subjects at particular values of X, there is MORE association
(Stronger).
Y
X
Correlation
To find the strength of the relationship
between two variables, we need
correlation.
The correlation is the standardized
slope it refers to the standard deviation
change in Y when you go up a standard
deviation in X.
Correlation
The correlation is the standardized slope it refers to the standard
deviation change in Y when you go up a standard deviation in X.
(X - X)
2

Recall that s.d. of x, Sx = n - 1
(Y - Y)
2

and the s.d. of y, Sy = n - 1
Sx
Pearson correlation, r = Sy b
Correlation
The Pearson Correlation, r:
tells the direction and strength of the
relationship between continuous variables
ranges from -1 to +1
is + when the relationship is positive and -
when the relationship is negative
the higher the absolute value of r, the stronger
the association
a standard deviation change in x corresponds
with r standard deviation change in Y
Correlation
The Pearson Correlation, r:
The pearson correlation is a statistic that is an
inferential statistic too.
r - (null = 0)
t
n-2
= (1-r
2
) (n-2)

When it is significant, there is a relationship in
the population that is not equal to zero!
Error Analysis
Y = a + bX This equation gives the conditional
mean of Y at any given value of X.





So In reality, our line gives us the expected mean of Y given each
value of X
The lines equation tells you how the mean on your dependent
variable changes as your independent variable goes up.
^
Y
^
X
Y
Error Analysis
As you know, every mean has a distribution around it--so
there is a standard deviation. This is true for conditional
means as well. So, you also have a conditional standard
deviation.



Conditional Standard Deviation or Root Mean Square Error
equals approximate average deviation from the line.
SSE ( Y - Y)
2

= n - 2 = n - 2
Y
^
X
Y
^
^
Error Analysis
The Assumption of Homoskedasticity:
The variation around the line is the same no matter the X.
The conditional standard deviation is for any given value of X.








If there is a relationship between X and Y, the conditional standard
deviation is going to be less than the standard deviation of Y--if this
is so, you have improved prediction of the mean value of Y by
looking at each level of X.
If there were no relationship, the conditional standard deviation
would be the same as the original, and the regression line would be
flat at the mean of Y.
Y
X
Y Conditional
standard
deviation
Original
standard
deviation
Error Analysis
So guess what?



We have a way to determine how much our
understanding of Y is improved when taking X
into accountit is based on the fact that
conditional standard deviations should be
smaller than Ys original standard deviation.
Error Analysis
Proportional Reduction in Error
Lets call the variation around the mean in Y Error 1.
Lets call the variation around the line when X is considered
Error 2.





But rather than going all the way to standard deviation to
determine error, lets just stop at the basic measure, Sum of
Squared Deviations.
Error 1 (E1) = (Y Y)
2 also called Sum of Squares

Error 2 (E2) = (Y Y)
2 also called Sum of Squared Errors
Y
X
Y Error 2 Error 1

R-Squared
Proportional Reduction in Error
To determine how much taking X into consideration reduces the
variation in Y (at each level of X) we can use a simple formula:
E1 E2 Which tells us the proportion or
E1 percentage of original error that
is Explained by X.





Error 1 (E1) = (Y Y)
2
Error 2 (E2) = (Y Y)
2
Y
X
Y
Error 2
Error 1

R-squared
r
2
= E1 - E2
E1

= TSS - SSE
TSS

= (Y Y)
2 -
(Y Y)
2

(Y Y)
2

r
2
is called the coefficient of
determination
It is also the square of the
Pearson correlation
Y
X
Y
Error 2
Error 1
R-Squared
R
2
:
Is the improvement obtained by using X (and drawing a line
through the conditional means) in getting as near as possible to
everybodys value for Y over just using the mean for Y alone.
Falls between 0 and 1
Of 1 means an exact fit (and there is no variation of scores
around the regression line)
Of 0 means no relationship (and as much scatter as in the
original Y variable and a flat regression line through the mean of
Y)
Would be the same for X regressed on Y as for Y regressed on
X
Can be interpreted as the percentage of variability in Y that is
explained by X.
Some people get hung up on maximizing R
2
, but this is too bad
because any effect is still a findinga small R
2
only indicates that
you havent told the whole (or much of the) story with your variable.
Error Analysis, SPSS
Some SPSS output (Anti- Gay Marriage regressed on Age):
r
2

(Y Y)2 - (Y Y)2
(Y Y)2


196.886 2853.286 = .069
Line to the Mean
Data points to the line
Data points to
the mean
Original SS
for Anti- Gay
Marriage
Error Analysis
Some SPSS output (Anti- Gay Marriage regressed on Age):
r
2

(Y Y)2 - (Y Y)2
(Y Y)2


196.886 2853.286 = .069
Line to the Mean
Data points to the line
Data points to
the mean
0 18 45 89
Age
Strong Oppose 5

Oppose 4

Neutral 3

Support 2

Strong Support 1
Anti- Gay
Marriage
M = 2.98
Colored lines are examples of:
Distance from each persons data point
to the line or modelnew, still
unexplained error.
Distance from line or model to Mean
for each personreduction in error.
Distance from each persons data point
to the Meanoriginal variables error.

ANOVA Table
X
Y
Mean
Q: Why do I see an ANOVA Table?
A: We bust up variance to get R
2
.
Each case has a value for distance
from the line (Y-bar
cond. Mean
) to Y-bar
big
,
and a value for distance from its Y
value and the line (Y-bar
cond. Mean
).
Squared distance from the line to the mean
(Regression SS) is equivalent to BSS, df =
1. In ANOVA, all in a group share Y-bar
group

The squared distance from the line to the
data values on Y (Residual SS) is
equivalent to WSS, df = n-2. In ANOVA, all in
a group share Y-bar
group

The ratio, Regression to Residual SS,
forms an F distribution in repeated
sampling. If F is significant, X explains
some variation in Y.
BSS
WSS
TSS
Line Intersects
Group Means
Dichotomous Variables
Y
X
0 1
1





2





3





4





5





6





7





8





9




1
0


Using a dichotomous independent variable,
the ANOVA table in bivariate regression will
have the same numbers and ANOVA results
as a one-way ANOVA table would (and
compare this with an independent samples t-
test).
^
0=Men: Mean = 6
1=Women: Mean = 4
Slope = -2 ; Y = 6 2X
Mean = 5
BSS
WSS
TSS
Regression, Inferential Statistics
Descriptive:
The equation for your line
is a descriptive statistic.
It tells you the real, best-
fitted line that minimizes
squared errors.

Inferential:
But what about the
population? What can we
say about the relationship
between your variables in
the population???
The inferential statistics
are estimates based on
the best-fitted line.

Recall that statistics are divided between descriptive
and inferential statistics.
Regression, Inferential Statistics
The significance of F, you already understand.





The ratio of Regression (line to the mean of Y) to Residual (line to
data point) Sums of Squares forms an F ratio in repeated sampling.
Null: r
2
= 0 in the population. If F exceeds critical F, then your
variables have a relationship in the population (X explains some of
the variation in Y).
Most extreme
5% of Fs
F = Regression SS / Residual SS
Regression, Inferential Statistics
What about the Slope or
Coefficient?
From sample to sample, different
slopes would be obtained.
The slope has a sampling
distribution that is normally
distributed.
So we can do a significance test.
-3 -2 -1 0 1 2 3 z

Regression, Inferential Statistics
Conducting a Test of Significance for the slope of the Regression Line

By slapping the sampling distribution for the slope over a guess of the
populations slope, H
o
, one determines whether a sample could
have been drawn from a population where the slope is equal H
o
.

1. Two-tailed significance test for -level = .05
2. Critical t = +/- 1.96
3. To find if there is a significant slope in the population,
H
o
:

= 0
H
a
:

0 ( Y Y )
2

4. Collect Data n - 2
5. Calculate t (z): t = b
o
s.e. =
s.e. ( X X )
2

6. Make decision about the null hypothesis
7. Find P-value

Correlation and Regression
Back to the SPSS output:
The standard error and t
appears on SPSS output
and the p-value too!
Correlation and Regression
Back to the SPSS output:
Y = 1.88 + .023X
So the GSS example, the
slope is significant. There is
evidence of a positive
relationship in the
population between Age and
Anti- Gay Marriage
sentiment. 6.9% of the
variation in Marriage
attitude is explained by age.
The older Americans get, the
more likely they are to
oppose gay marriage.

A one year increase in age elevates anti attitudes by .023 scale units. There is a weak
positive correlation. A s.d, increase in age produces a .023 s.d. increase in anti scale units.