Relationship between two variables e.g, as education , what does income do? Scatterplot Bivariate Methods.

• Relationship between two variables

• e.g, as education , what does income do?

• Scatterplot

Bivariate Methods

Correlation

Linear Correlation

Source: Earickson, RJ, and Harlin, JM. 1994. Geographic Measurement and Quantitative Analysis. USA: Macmillan College Publishing Co., p. 209.

0.0

0.1

0.2

0.3

0.4

0.5

0.6

4 5 6 7 8 9 10 11 12 13

TMI

Th

eta

0.0

0.1

0.2

0.3

0.4

0.5

0.6

2.5 3.0 3.5 4.0 4.5 5.0 5.5

TMI

Th

eta

0.0

0.1

0.2

0.3

0.4

0.5

0.6

4 5 6 7 8 9 10 11 12 13

TMI

Th

eta

0.0

0.1

0.2

0.3

0.4

0.5

0.6

2.5 3.0 3.5 4.0 4.5 5.0 5.5

TMI

Th

eta

0.0

0.1

0.2

0.3

0.4

0.5

0.6

4 5 6 7 8 9 10 11 12 13

TMI

Th

eta

0.0

0.1

0.2

0.3

0.4

0.5

0.6

2.5 3.0 3.5 4.0 4.5 5.0 5.5

TMIT

het

a

Wet – May 29/30 Avg. – June 26/28 Dry – August 22

Pond B

ranch - P

G 11.25m

DE

MG

lyndon – LID

AR

0.5m

DE

M 11x11

R2=0.71

R2=0.29

R2=0.79

R2=0.24

R2=0.79

R2=0.10

Theta-TVDI ScatterplotsGlyndon Field Sampled Soil Moisture

versus TVDI from a 3x3 kernel

0.00

0.05

0.10

0.15

0.20

0.25

0.30

0.35

0.40

0.45

0.50

0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0

TVDI (3x3 kernel)

Vo

lum

etri

c S

oil

Mo

istu

rePond Branch Field Sampled Soil Moisture

versus TVDI from a 3x3 kernel

0.00

0.05

0.10

0.15

0.20

0.25

0.30

0.35

0.40

0.45

0.50

0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0

TVDI (3x3 kernel)

Vo

lum

etr

ic S

oil

Mo

istu

re

Oxford Tobacco Research Station Field Sampled Soil Moisture versus TVDI from a 3x3 kernel

0.00

0.05

0.10

0.15

0.20

0.25

0.30

0.35

0.40

0.45

0.50

0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0

TVDI (3x3 kernel)

Vo

lum

etr

ic S

oil

Mo

istu

re

API-TVDI Scatterplot

Covariance: Interpreting Scatterplots

• General sense (direction and strength)

• Subjective judgment

• More objective approach

• Extent to which variables Y and X vary together

• Covariance

Covariance Formulae

Cov [X, Y] = (xi - x)(yi - y)i=1

i=n1

n - 1

Covariance Example

Glyndon Field Sampled Soil Moisture versus TVDI from a 3x3 kernel

0.00

0.05

0.10

0.15

0.20

0.25

0.30

0.35

0.40

0.45

0.50

0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0

TVDI (3x3 kernel)

Vo

lum

etri

c S

oil

Mo

istu

re

TVDISoil

Moisture

0.274 0.4140.542 0.3590.419 0.3960.286 0.4580.374 0.3500.489 0.3570.623 0.2550.506 0.1890.768 0.1710.725 0.119

Covariance Example

TVDI (x)Soil

Moisture (y)

(x - xbar) (y - ybar)(x - xbar) * (y - ybar)

0.274 0.414 -0.227 0.107 -0.0243050.542 0.359 0.042 0.052 0.00216240.419 0.396 -0.082 0.090 -0.0073230.286 0.458 -0.215 0.151 -0.0324240.374 0.350 -0.127 0.044 -0.0055530.489 0.357 -0.011 0.050 -0.0005660.623 0.255 0.122 -0.052 -0.0063740.506 0.189 0.005 -0.118 -0.0006180.768 0.171 0.267 -0.136 -0.0362820.725 0.119 0.225 -0.188 -0.042289

Mean 0.501 0.307 -0.15357-0.017063

SumCovariance

1

2 3

45

How Does Covariance Work?

• X and Y are positively related

• xi > x yi > y

• xi < x yi < y

• X and Y are negatively related

• xi > x yi < y

• xi < x yi > y

__ __

__ __

__ __

__ __

Interpreting Covariances

• Direction & magnitude

• Cov[X,Y] > 0 positive

• Cov[X, Y] < 0 negative

• abs(Cov[X, Y]) ↑ strength ↑

• Magnitude ~ units

Covariance Correlation

• Magnitude ~ units

• Multiple pairs of variables not comparable

• Standardized covariance

• Compare one such measure to another

Pearson’s product-moment correlation coefficient

Cov [X, Y]

sXsY

r =

r (xi - x)(yi - y)i=1

i=n

(n - 1) sXsY

=

ZxZyr i=1

i=n

(n - 1)=

Pearson’s Correlation Coefficient

• r [–1, +1]

• abs(r) ↑ strength ↑

• r cannot be interpreted proportionally

• ranges for interpreting r values 0 - 0.2 Negligible

0.2 - 0.4 Weak

0.4 - 0.6 Moderate

0.6 - 0.8 Strong

0.8 - 1.0 Very strong

Example

• X = TVDI, Y = Soil Moisture

• Cov[X, Y] = -0.017063

• SX = 0.170, SY = 0.115

• r ?

Glyndon Field Sampled Soil Moisture versus TVDI from a 3x3 kernel

0.00

0.05

0.10

0.15

0.20

0.25

0.30

0.35

0.40

0.45

0.50

0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0

TVDI (3x3 kernel)

Vo

lum

etri

c S

oil

Mo

istu

re

Pearson’s r - Assumptions

1. interval or ratio

2. Selected randomly

3. Linear

4. Joint bivariate normal distribution

Interpreting Correlation Coefficients

• Correlation is not the same as causation!

• Correlation suggests an association between

variables

1. Both X and Y are influenced by Z

Interpreting Correlation Coefficients

2. Causative chain (i.e. X A B Y)

e.g. rainfall soil moisture ground water runoff

3. Mutual relationship

e.g., income & social status

4. Spurious relationship

e.g., Temperature (different units)

5. A true causal relationship (X Y)

Interpreting Correlation Coefficients

6. A result of chance

e.g., your annual income vs. annual population of the world

Interpreting Correlation Coefficients

7. Outliers

(Source: Fang et al., 2001, Science, p1723a)

Interpreting Correlation Coefficients

• Lack of independence

– Social data

– Geographic data

– Spatial autocorrelation

A Significance Test for r

• An estimator

r

= 0 ?

• t-test

A Significance Test for r

ttest = r

SEr

=r

1 - r2

n - 2

=r n - 2

1 - r2

df = n - 2

A Significance Test for r

H0: = 0

HA: 0

ttest = r n - 2

1 - r2