Analyzing Road Accident Data: A Statistical Report on Four Countries

This report presents a statistical analysis and interpretation of road accident data from 1994 to 2016 for Russia, Canada, Japan, and Australia, sourced from the OECD. The analysis includes measures of central tendency (mean, median, mode) and dispersion (standard deviation, variance, minimum, maximum, range, skewness). Russia exhibits the highest average annual road accident-related deaths and data spread, while Japan records the lowest. Correlation analysis reveals positive correlations among all countries, with strong correlations between Canada, Australia, and Japan. A hypothesis test comparing the means of Canada and Australia indicates no significant difference. The report concludes that while all four countries show some correlation in road accident trends, Russia's figures are significantly higher than the other three, and Canada and Australia have statistically similar road accident death rates.

Road Accident Statistics
Russia, Canada, Japan, and Australia

Road Accident Data
• The presentation will cover statistical analysis and
interpretation of data relating to the number of
annual deaths in four countries that can be
attributed to road accidents. The Data used is for the
time period between 1994 and 2016
• The Data was retrieved from the online data archives
of the Organization for Economic Co-operation and
Development (OCED). Data was collected for several
nations however we will concentrate on Russia,
Canada, Japan, and Australia.

Measures of Center Tendency
Russia Canada Australia Japan
Mean 205.62 82.73 77.92 64.19
Median 201.87 87.91 78.65 62.59
Mode 201.87
Russia has the highest number of average annual road accident
related deaths, with the other three ranking 2nd to 4th in the
following order Canada, Australia, and finally Japan. The ranking
is the same in terms of the median deaths observed between
1994 and 2016. Lastly, only Russia has a mode which is also
similar to the median; this indicates that median is a better
estimate of data center than mean (Pyne, Rao, & Rao, 2016).

Measures of Dispersion
Russia Canada Australia Japan
Standard
Deviation 26.27 19.28 19.58 21.87
Variance 690.02 371.57 383.36 478.27
Minimum 140.69 51.83 49.02 36.99
Maximum 246.13 112.86 111.61 102.21
Range 105.44 61.03 62.59 65.22
Skewness -0.47 -0.26 0.07 0.29

Interpreting Measures of Variability
• The standard deviation indicates the spread of individual
data points for the mean. While, Variance measures the
overall dispersion of data points from the mean
(Vartanian, 2010). As such, the most spread out data set
has to be that of Russia, and the least spread out data set
belongs to Canada.
• The highest annual road accident related deaths figure for
all countries for the period 1994 -2016 was recorded by
Russia ( 246.13) and the lowest was observed in Japan
(36.99) . None of the data for the four countries is
normally distributed; As such, the data for Russia and
Canada is negatively skewed, and data for Australia and
Japan is positively skewed.

Line Graph for All Four Countries
1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016
0
50
100
150
200
250
300
Annual Road Accident Related Deaths
Russia
Canada
Australia
Japan
Year
Number of Deaths

Interpreting Line Graph
• It is clear that Russia recorded the highest road
accident related deaths between 1994 and 2016.
Moreover, none of the other three come close to
recording similar annual mortality figures.
• Canada and Australia recorded roughly the same
annual road accident deaths. Lastly, Japan recorded
the lowest death figures throughout the assessment
period. It is important to note that Canada’s data is
provided for period between 1994 and 2015; As
such, 2016 data is missing.

Correlation Analysis
Russia Canada Australia Japan
Russia 1
Canada 0.547036 1
Australia 0.456553 0.98266 1
Japan 0.388248 0.956342 0.97624 1
•All four countries’ datasets have positive correlations with each
other; this means an increment in road accident deaths in one
countries will also be witnessed in other (although not in the same
magnitude) (EMC, 2015).
•However, Canada, Australia, and Japan have very strong
correlations with each other. But the strongest correlation is
between Canada and Australia. Russia has a moderate, fair, and
weak correlation with Canada, Australia and Japan respectively.

Hypothesis for Sample Means
t-Test: Two-Sample Assuming Unequal Variances
Canada Australia
Mean 82.73091 77.92
Variance 371.5749 383.3564273
Observations 22 23
Hypothesized
Mean Difference 0
df 43
t Stat 0.830487
P(T<=t) one-tail 0.205426
t Critical one-tail 1.681071
P(T<=t) two-tail 0.410851
t Critical two-tail 2.016692

Interpreting Hypothesis Results
• We suspect that the dataset for Australia and Canada were quite
similar. As such, a hypothesis test is conducted to show;
• Null Hypothesis H0: there is no significant difference between the
means for Canada and Australia i.e. U1=U2
• Alternative hypothesis H1: there is no significant difference
between the means for Canada and Australia i.e. U1≠U2
• A t-test was used at 5% significance level because the samples
are small (Holcomb, 2016). We will not reject the null hypothesis
because the t-score is within the lower and upper critical values
i.e. -2.016< 0.83<2.016. As such, there is no significant difference
between the means of Canada and Australia (road accident data).

References
• EMC, E. S. (2015). Data Science and Big Data
Analytics: Discovering, Analyzing, Visualizing and
Presenting Data. Hoboken: John Wiley & Sons.
• Holcomb, Z. C. (2016). Fundamentals of Descriptive
Statistics. Didcot: Taylor & Francis.
• Pyne, S., Rao, B. P., & Rao, S. (2016). Big Data
Analytics: Methods and Applications. Berlin: Springer.
• Vartanian, T. P. (2010). Secondary Data Analysis.
Oxford: Oxford University Press.