Вы находитесь на странице: 1из 6

Advanced Statistical Methods (BAZG524) Assignment

U.S. Department of
Transportation

GROUP 4

2018MC21003 Swastik Mishra


2018MC21084 C.J. Narasimhan
2018MC21029 Akshay Upendran
2018MC21013 Pankti Dave Yogesh
2018MC21078 Shivangi Vajpai
2018MC21082 Subhashish Sahu
U.S. Department of Transportation
As part of a study on transportation safety, the U.S. Department of Transportation collected data on the number of
fatal accidents per 1000 licenses and the percentage of licensed drivers under the age of 21 in a sample of 42 cities.
Data collected over a one-year period follow. These data are tabulated below.

Percent Under 21 Fatal Accidents per 1000 Licenses


13 2.962
12 0.708
8 0.885
12 1.652
11 2.091
17 2.627
18 3.83
8 0.368
13 1.142
8 0.645
9 1.028
16 2.801
12 1.405
9 1.433
10 0.039
9 0.338
11 1.849
12 2.246
14 2.855
14 2.352
11 1.294
17 4.1
8 2.19
16 3.623
15 2.623
9 0.835
8 0.82
14 2.89
8 1.267
15 3.224
10 1.014
10 0.493
14 1.443
18 3.614
10 1.926
14 1.643
16 2.943
12 1.913
15 2.814
13 2.634
9 0.926
17 3.256
1. Develop numerical and graphical summaries of the data.

Doing a quick scatter plot reveals that there is indeed relationship between two variables.

Fatal Accidents per 1000 Licenses


4.5
4
3.5
3
2.5
2
1.5
1
0.5
0
0 2 4 6 8 10 12 14 16 18 20

Both variables don’t follow normal distribution as evident from the below histograms

Histogram Histogram
8 7
7 6
6
5
Frequency

Frequency

5
4 4
3 3
2 2
1 1
0 0

Fatal Accidents per 1000 Licenses Percent Under 21


Below is the statistical summary of the data. We can see that variation of percent of licensed driver under age 21 is
much more than fatal accidents per 1000 licenses.

Percent Under 21 Value Fatal Accidents per 1000 Licenses Value

Mean 12.26 Mean 1.92


Standard Error 0.48 Standard Error 0.17
Median 12 Median 1.881
Mode 8 Mode #N/A
Standard Deviation 3.13 Standard Deviation 1.07
Sample Variance 9.81 Sample Variance 1.15
Kurtosis -1.14 Kurtosis -0.97
Skewness 0.21 Skewness 0.19
Range 10 Range 4.061
Minimum 8 Minimum 0.039
Maximum 18 Maximum 4.1
Sum 515 Sum 80.74
Count 42 Count 42
Confidence Level(95.0%) 0.9759 Confidence Level(95.0%) 0.3337

2. Use regression analysis to investigate the relationship between the


number of fatal accidents and the percentage of drivers under the age of
21. Discuss your findings.

We have done linear regression fitting of the data, below are the output.

Regression Statistics Value


Multiple R 0.83938748
R Square 0.704571341
Adjusted R Square 0.697185624
Standard Error 0.589350288
Observations 42

We can see from the R Square value that the linear regression model explains 70% of the variation in fatal accidents
through drivers under age of 21.
Below are ANOVA tables for the model

df SS MS F Significance F

Regression 1 33.134 33.134 95.396 0.00000


Residual 40 13.893 0.347
Total 41 47.028

Standard P- Lower Upper Lower Upper


Coefficients Error t Stat value 95% 95% 95.0% 95.0%
Intercept -1.5974 0.3717 -4.2979 0.0001 -2.3486 -0.8462 -2.3486 -0.8462
Percent Under 21 0.2871 0.0294 9.7671 0.0000 0.2277 0.3465 0.2277 0.3465

We can see that p-value of F-test is significant. Also p-values of parameters are significant.

Percent Under 21 Line Fit Plot


5
Fatal Accidents per 1000

4
3
Licenses

2
1
0
0 5 10 15 20
Percent Under 21

Series1 Predicted Fatal Accidents per 1000 Licenses

Percent Under 21 Residual Plot


2
1.5
1
Residuals

0.5
0
-0.5 0 5 10 15 20
-1
-1.5
Percent Under 21

The line fit plot shows that predicted values are close to actual values and also the relationship that with increase in
percentage of licensed drivers under the age of 21 number of fatal accidents also increases.

Residual plot shows no pattern so we can conclude that linear regression is the best fit for modelling this problem.

We reject the null hypothesis that two variables are independent and conclude that there is relationship between
number of fatal accidents and percentage of licensed drivers under the age of 21.
3. What conclusion and recommendations can you derive from your
analysis?
We saw from regression analysis of the given data that number of fatal accidents increases with the increase in
percentage of licensed drivers under the age of 21. It can be recommended to increase the minimum age for
applying driver’s license to increase transportation safety.

Вам также может понравиться