Practical 7 – CP2403

(Please ensure you show your work to your tutor once completed as each practical is 3
marks.)

Part 1 – Download the Jupyter notebook for Module 7 and run the notebook

Part 2
Download the Jupyter Notebook Template for Prac 7 from LearnJCU. Complete the template
& run the code. Refer to Module 7 Lecture Jupyter Notebook for help
Complete the questions in Part 3 as you work on the Prac 7 template

Part 3
Scenario 1
We want to find out the association between electricity use per person (relectricperperson (x))
and oil use per person (oilperperson (y))
Step 1: hypothesis
Null hypothesis (Ho):
There is more person who use electricity than oil
Or there is a negative correlation between electricity use per person and oil use per person.

Step 2: Data Selection

- 2010 oil Consumption per capital (tons per year and person)
- 2008 residential electricity consumption, per person (kWh)

Step 3: Assess the evidence

Scatter Plot
Pearson correlation (r):
0.5249373779159885, 1.0020621767836594e-05

As r is positive, there is a positive relationship. Also, r is close to 1, then there is a strong

positive relationship.

p: 1.00e-05 = -2.28

since p is less than 0.05, will indicate that the correlation will hold for larger population.

Step 4: Draw Conclusion

Accept (There is a strong positive correlation between electricity use per person and oil use
per person.)
Scenario 2
We want to find out the association between CO2 emission (co2emissions (x)) and oil use per
person (oilperperson (y))
Step 1: hypothesis
Null hypothesis (Ho):
There is more amount of oil use per person than CO2 emission.
Or there is a negative correlation between oil use per person and CO2 emission.

Step 2: Data Selection

- 2010 oil Consumption per capital (tons per year and person)
- 2006 cumulative CO2 emission (metric tons), Total amount of CO2 emission in
metric tons since 1751.

Step 3: Assess the evidence

Scatter Plot

Pearson correlation (r):

0.04444201231228792, 0.7294518840123033
As r is positive, there is a positive relationship. Also, r is close to 1, then there is a strong
positive relationship.

p: 0.729
As p is greater than 0.05, will indicate that the correlation will hold for smaller population.

Step 4: Draw Conclusion

Accept (There is a strong positive correlation between oil use per person and CO2
emmision.)
Scenario 3
We want to perform regression analysis between electricity use per person (relectricperperson
(x)) and oil use per person (oilperperson (y))
1: Scatter plot with regression line

2: Regression Analysis results

OLS regression model for the association between Electric Use Per Person and Oil Per Person
OLS Regression Results
==============================================================================
Dep. Variable: relectricperperson R-squared: 0.276
Model: OLS Adj. R-squared: 0.264
Method: Least Squares F-statistic: 23.20
Date: Mon, 17 Dec 2018 Prob (F-statistic): 1.00e-05
Time: 19:27:29 Log-Likelihood: -559.23
No. Observations: 63 AIC: 1122.
Df Residuals: 61 BIC: 1127.
Df Model: 1
Covariance Type: nonrobust
================================================================================
coef std err t P>|t| [0.025 0.975]
--------------------------------------------------------------------------------
Intercept 860.2499 286.936 2.998 0.004 286.485 1434.015
oilperperson 590.3956 122.567 4.817 0.000 345.309 835.483
==============================================================================
Omnibus: 35.040 Durbin-Watson: 1.908
Prob(Omnibus): 0.000 Jarque-Bera (JB): 211.745
Skew: 1.246 Prob(JB): 1.05e-46
Kurtosis: 11.629 Cond. No. 3.27
==============================================================================

3: Regression line – if valid

oilperperson = b + m(relectricperperson)
oilperperson = 860.2499+590.3956*(relectricperperson)

4: Residual plot – if required

5: Conclusion from residual plot – if valid

- Assume relationship is linear
- Residuals are normally distributed
Scenario 4
We want to perform regression analysis between CO2 emission (co2emissions (x)) and oil
use per person (oilperperson (y))
1: Scatter plot with regression line

2: Regression Analysis results

OLS regression model for the association between CO2 emission and Oil Use Per Person
OLS Regression Results
==============================================================================
Dep. Variable: co2emissions R-squared: 0.002
Model: OLS Adj. R-squared: -0.014
Method: Least Squares F-statistic: 0.1207
Date: Mon, 17 Dec 2018 Prob (F-statistic): 0.729
Time: 19:27:30 Log-Likelihood: -1633.3
No. Observations: 63 AIC: 3271.
Df Residuals: 61 BIC: 3275.
Df Model: 1
Covariance Type: nonrobust
================================================================================
coef std err t P>|t| [0.025 0.975]
--------------------------------------------------------------------------------
Intercept 1.368e+10 7.28e+09 1.880 0.065 -8.71e+08 2.82e+10
oilperperson 1.08e+09 3.11e+09 0.347 0.729 -5.14e+09 7.3e+09
==============================================================================
Omnibus: 119.379 Durbin-Watson: 1.743
Prob(Omnibus): 0.000 Jarque-Bera (JB): 4709.059
Skew: 6.136 Prob(JB): 0.00
Kurtosis: 43.538 Cond. No. 3.27
==============================================================================

Warnings:
[1] Standard Errors assume that the covariance matrix of the errors is correctly specified.
3: Regression line – if valid
oilperperson = b + m(co2emissions)
oilperperson = 1.368e+(1.08e=09) * co2emissions

4: Residual plot – if required

5: Conclusion from residual plot – if valid

- Residuals are normally distributed
Scenario 5
We want to perform regression analysis between electricity use per person (relectricperperson
(x)) and oil use per person (oilperperson (y)) for LOW CO2 emission countries
1: Scatter plot with regression line

2: Regression Analysis results

OLS regression model for the association between Electricty Use Per Person and Oil Use Per Person for
LOW CO2 Emission countries
OLS Regression Results
==============================================================================
Dep. Variable: relectricperperson R-squared: 0.244
Model: OLS Adj. R-squared: 0.194
Method: Least Squares F-statistic: 4.840
Date: Mon, 17 Dec 2018 Prob (F-statistic): 0.0439
Time: 19:27:43 Log-Likelihood: -151.33
No. Observations: 17 AIC: 306.7
Df Residuals: 15 BIC: 308.3
Df Model: 1
Covariance Type: nonrobust
================================================================================
coef std err t P>|t| [0.025 0.975]
--------------------------------------------------------------------------------
Intercept 917.9271 557.225 1.647 0.120 -269.770 2105.625
oilperperson 337.1159 153.240 2.200 0.044 10.492 663.739
==============================================================================
Omnibus: 11.718 Durbin-Watson: 2.325
Prob(Omnibus): 0.003 Jarque-Bera (JB): 9.464
Skew: 1.203 Prob(JB): 0.00881
Kurtosis: 5.751 Cond. No. 4.53
==============================================================================
3: Regression line
oilperperson = b + m(co2emission)
oilperperson=917.9271 + 337.1159 (relectricperperson)

4: Residual plot – if required

5: Conclusion from residual plot – if valid

Residuals are normally distributed

Scenario 6
We want to perform regression analysis between electricity use per person (relectricperperson
(x)) and oil use per person (oilperperson (y)) for MEDIUM CO2 emission countries
1: Scatter plot with regression line

2: Regression Analysis results

OLS regression model for the association between Electricty Use Per Person and Oil Use Per Person for
MEDIUM CO2 Emission countries
OLS Regression Results
==============================================================================
Dep. Variable: relectricperperson R-squared: 0.619
Model: OLS Adj. R-squared: 0.597
Method: Least Squares F-statistic: 27.61
Date: Mon, 17 Dec 2018 Prob (F-statistic): 6.45e-05
Time: 19:27:52 Log-Likelihood: -153.72
No. Observations: 19 AIC: 311.4
Df Residuals: 17 BIC: 313.3
Df Model: 1
Covariance Type: nonrobust
================================================================================
coef std err t P>|t| [0.025 0.975]
--------------------------------------------------------------------------------
Intercept -55.0934 371.418 -0.148 0.884 -838.716 728.529
oilperperson 1209.1409 230.096 5.255 0.000 723.680 1694.602
==============================================================================
Omnibus: 4.056 Durbin-Watson: 1.590
Prob(Omnibus): 0.132 Jarque-Bera (JB): 1.998
Skew: -0.690 Prob(JB): 0.368
Kurtosis: 3.785 Cond. No. 4.09
3: Regression line
oilperperson = b + m(co2emissions)
oilperperson = -55.0934+1209.1409(co2emissions)

4: Residual plot – if required

5: Conclusion from residual plot – if valid

Residuals are normally distributed
Scenario 7
We want to perform regression analysis between electricity use per person (relectricperperson
(x)) and oil use per person (oilperperson (y)) for HIGH CO2 emission countries
1: Scatter plot with regression line

2: Regression Analysis results

OLS regression model for the association between Electricty Use Per Person and Oil Use Per Person for
HIGH CO2 Emission countries
OLS Regression Results
==============================================================================
Dep. Variable: relectricperperson R-squared: 0.619
Model: OLS Adj. R-squared: 0.597
Method: Least Squares F-statistic: 27.61
Date: Mon, 17 Dec 2018 Prob (F-statistic): 6.45e-05
Time: 19:30:38 Log-Likelihood: -153.72
No. Observations: 19 AIC: 311.4
Df Residuals: 17 BIC: 313.3
Df Model: 1
Covariance Type: nonrobust
================================================================================
coef std err t P>|t| [0.025 0.975]
--------------------------------------------------------------------------------
Intercept -55.0934 371.418 -0.148 0.884 -838.716 728.529
oilperperson 1209.1409 230.096 5.255 0.000 723.680 1694.602
==============================================================================
Omnibus: 4.056 Durbin-Watson: 1.590
Prob(Omnibus): 0.132 Jarque-Bera (JB): 1.998
Skew: -0.690 Prob(JB): 0.368
Kurtosis: 3.785 Cond. No. 4.09

3: Regression line
oilperperson = b + m(co2emissions)
oilperperson = -55.0934+1209.1409(co2emissions)

4: Residual plot – if required

5: Conclusion from residual plot – if valid

Residuals are normally distributed