A description of the data:
• who collected it,
• what variables it contains and
• summary statistics for the dataset.
Table 1: Summary statistics
Statistic N Mean St. Dev. Min Max
num_drivers 51 15.790 4.122 5.900 23.900
perc_speeding 51 31.725 9.633 13 54
perc_alcohol 51 30.686 5.132 16 44
perc_not_distracted 51 85.922 15.159 10 100
perc_no_previous 51 88.725 6.960 76 100
insurance_premiums 51 886.958 178.296 641.960 1,301.520
losses 51 134.493 24.836 82.750 194.780
3 Data analysis
3.1 Overview
Some relevant plots, with brief discussions of the plots.
1
10 15 20
700
900
1100
1300
Number of drivers involved in fatal collisions per billion miles
Car insur
ance premiums ($)
2
80 100 120 140 160 180
700
900
1100
1300
Losses incurred by insurance companies for collisions per insured driver ($)
Car insur
ance premiums ($)
3.2 In-depth analysis
A more in-depth statistical analysis, with a brief discussion of what you find:
• what variables matter
• what quantitative effects did you find
3
Table 2: Regression results
Dependent variable:
insurance_premiums
(1) (2) (3)
losses 4.473∗∗∗ 4.427∗∗∗ 4.485∗∗∗
(0.802) (0.790) (0.804)
num_drivers −7.677 −8.065
(4.759) (4.944)
perc_speeding 1.103
(2.165)
perc_alcohol 1.914
(4.150)
Constant 285.325∗∗ 412.724∗∗∗ 317.442∗
(109.669) (133.726) (182.780)
Observations 51 51 51
R2 0.388 0.420 0.428
Adjusted R2 0.376 0.396 0.378
Residual Std. Error 140.866 (df = 49) 138.618 (df = 48) 140.581 (df = 46)
F Statistic 31.101∗∗∗ (df = 1; 49) 17.361∗∗∗ (df = 2; 48) 8.607∗∗∗ (df = 4; 46)
Note: ∗p<0.1; ∗∗p<0.05; ∗∗∗p<0.01
4 Discussion and conclusion
A discussion of your results.
• What economic explanation – in terms of inventives – can you find?
• Discussion of causality.
• Why might we doubt the conclusions?