Analysis of Environmental Factors on Bike Sharing Program in DC

This report analyzes the impact of environmental factors on bike sharing programs in Washington, D.C. The study examines the influence of temperature, humidity, and wind speed on bike usage, utilizing data from the Capital Bikeshare system from 2011 to 2012. The methodology includes descriptive statistics, normality tests, and correlation analysis to assess the relationships between environmental variables and bike rental counts. The report discusses the quality of the sample data, sampling errors, and statistical techniques such as correlation and covariance. The findings suggest a positive association between temperature and bike sharing, while humidity and wind speed show inverse relationships. The report concludes with an overview of the impact of environmental factors on bike sharing programs and their implications, with a focus on policy and user behavior.

Running head: IMPACT OF ENVIRONMENTAL FACTOR ON BIKE SHARING PROGRAM
IN WASHINGTON, D.C.
Impact of Environmental Factor on Bike Sharing Program in Washington, D.C.
Name of the Student
Name of the University
Course ID

1
IMPACT OF ENVIRONMENTAL FACTOR ON BIKE SHARING PROGRAM IN
WASHINGTON, D.C.
Abstract
There is a rapid growth of bicycle renting program in the last few years worldwide. Cities
located in different countries have implemented the system of bicycle renting. In USA,
Washington D.C is considered to have one of the largest bikesharing system. However,
unfavorable environmental condition especially unfavorable weather situation is one obstacle in
promoting bicycle sharing program. In this connection, the paper attempts to make an analysis on
impact of different environmental factor on bike sharing program in Washington, D.C. For this
purpose relevant data have been collected analyzed with the use of appropriate statistical
techniques. From the analysis, temperature has found to have a positive association with bike
sharing program. Humidity and Wind speed are inversely associated with bike sharing program.

2
IMPACT OF ENVIRONMENTAL FACTOR ON BIKE SHARING PROGRAM IN
WASHINGTON, D.C.
Table of Contents
Introduction......................................................................................................................................3
Methodology....................................................................................................................................5
Statistical Data Analysis..................................................................................................................6
Sample data..................................................................................................................................6
Quality of sample data.................................................................................................................9
Sampling Error...........................................................................................................................12
Correlation.................................................................................................................................19
Covariance.................................................................................................................................24
Conclusion.....................................................................................................................................26
References......................................................................................................................................28

3
IMPACT OF ENVIRONMENTAL FACTOR ON BIKE SHARING PROGRAM IN
WASHINGTON, D.C.
Introduction
The current bike sharing programs are totally different from traditional method of bike
sharing as modern technology has involved in the overall process that highlights the
effectiveness of the bike sharing programs. At the initial level, modern bike sharing programs are
convenient, cheap and flexible for the people. The bikes are required to be stored at the station
from where the people can access the bike and reach their destination (El-Assi, Mahmoud and
Habib 2017). Sharing is also another good option as it would help in reducing the cost as well as
the emission which also provide positive point on environmental health. It is associated with the
whole process and from the payment of membership fees which automatically gets deducted
every time the user uses the bike sharing facility (Wang et al. 2015). The system is made easy for
the users which made them feel comfortable to access the bikes at their desired locations. This
particular overall procedure mainly tracked by the application which is specially designed for the
bike sharing program. The role of traffic, environmental issues and health issues has played a
vital role for gaining the interest in the mind of the users (Tran, Ovtracht and D’arcier 2015). The
attractiveness for the research is to oppose the transport services which includes the buses and
taxis. The application mainly tracked down the bike along with the location through which it
travels. The virtual sensor mainly turns on for locating the bike along with collecting the data of
the overall journey. The major impact that has put on the life of the people and also on the
environment are significantly important for the overall process.
The popularity of bike sharing mainly provides different socio-economic benefits as well
as environmental benefits which mainly includes the overall emission of carbon dioxide and
other harmful gases that might led to serious health condition (Mateo-Babiano, Kumar and Mejia
2017). The reduction in noise pollution along with the air pollution is also another benefit that

4
IMPACT OF ENVIRONMENTAL FACTOR ON BIKE SHARING PROGRAM IN
WASHINGTON, D.C.
would be enjoyed with the process of bike sharing (Zhao 2014). Reduction in the overall traffic
congestion is also another factors that is enhanced by the process of bike sharing program which
would lead to reduction in overall level of pollution in the environment. Moreover, the cost that
are generally incurred by the individuals for travelling from one destination to another would
also get reduced due to the bike sharing program (Ricci 2015). This would help in increasing the
individual income and decreasing the overall expenses that are generally incurred by the people.
It has also been assumed that bikes are only used for replacing the taxis and other cabs along
with the buses which makes too much pollution towards the environment and emits several
harmful gases (Zhang et al. 2015). The overall health condition of humans has also degraded
with increase in number of vehicle along with their emission and that has to be reduced at a huge
rate (Fishman, Washington and Haworth 2014). Bike sharing program is the types of program
which takes a small step towards the reduction of the emission along with increasing the overall
health of humans as well as the humans.
The environmental factors are also required to be categorized on the level of efficiency of
bike sharing programs and that have to be considered with the overall literature of the research. It
also acts as an active mode of transportation as the people do not have to wait for the public
transport for travelling from one location to another location (Handy, Van Wee and Kroesen
2014). Another important aspect is that it helps in saving the fuel as it already faces limitations in
current environmental condition. Reduction in usage of petrol or diesel also assist in reducing the
overall cost and pollution that affects the most for the environment (Chen 205). Weather
condition is also highly correlated with the bike sharing programs as it effects the overall sharing
behaviors of the bike. This particular study mainly points out the research that is required to be
conducted for analyzing the impact of bike sharing program on bike sharing programs (Ji et al.

5
IMPACT OF ENVIRONMENTAL FACTOR ON BIKE SHARING PROGRAM IN
WASHINGTON, D.C.
2014). The regression analysis is to be performed which help in prediction of bike rental count
based on hourly and daily based on the environmental as well as seasonal setting (Faghih-Imani
et al. 2014). The condition that are required to be analyzed highlights the event and anomaly
detection which assist in counting of bikes that are correlated to some events in the towns and
that can also be traceable with specific search engines.
The impact of environmental factors on the bike sharing programs has put severe effect
on the environment along with increasing the health benefits. The policy makers mainly
encourage people to use the bike on a basis of sharing which would help in reducing health
effects along with certain health benefits. It also shed light on the convenience and the behaviors
which would provide safety measures that attracts most of the people to use bike sharing
programs.
Methodology
In order to analyze the impact of environmental factor on bike sharing program in
Washington data are collected on different variables. The core data set related to bike sharing is
collected from Capital Bike share system of Washington D.C. The information related to weather
condition and trips in different days of year are added to the data of bike-sharing. Day to Day are
collected for the two consecutive years ranged from 2011 to 2012. To examine demand of bike
sharing program in different days of the year days are classified as holiday, weekday or working
day. State of weather in different days are modelled by different weather condition such as clear,
few clouds, partly cloudy, Mist, Cloudy, broken clouds, light snow, light rain, thunderstorm,
scattered clouds, heavy rain, snow, fog and such other. Different environmental factor that are
included in the dataset are temperature, humidity and wind speed. The method of data analysis
first includes description of different variables in the data set. Before using the sample data for

6
IMPACT OF ENVIRONMENTAL FACTOR ON BIKE SHARING PROGRAM IN
WASHINGTON, D.C.
statistical analysis quality of the data needs to be examined. For examining quality of the data
normality test has been performed along with estimation of sampling error. Descriptive statistical
measures such as mean, median, standard deviation, coefficient of variation and others are
estimated to make an overall assessment of the data. Finally, correlation coefficient and
covariance are estimated to obtain impact of different factors on environment.
Statistical Data Analysis
Sample data
For obtaining a robust result, sample data should be checked for normality. In order to
examine nature of the statistical distribution of the data set, each of the variables are plotted. The
statistical analysis includes iteration process for estimating the parameters choosing appropriate
Probability Distribution Function. The examination of normality distribution is made by plotting
the PDF of the sample data. Table 1 below presents description of variables included in the
sample.
Table 1: Description of parameters of the sample data
Sl.No Variable Name Description Average
1. Instant Record Index -
2. dteday Date -
3. holiday Weather the day is holiday or not -
4. Weekday Day of the week -
5. Workingday If the day is neither weekend nor
holiday is 1, otherwise 0
-
6. Weathersit Different condition of weather -

7
IMPACT OF ENVIRONMENTAL FACTOR ON BIKE SHARING PROGRAM IN
WASHINGTON, D.C.
7. temp Normalized temperature in Celsius 0.50
8. hum Normalized humidity 0.63
9. windspeed Normalized wind speed 0.19
10. cnt Count of total rental bikes 4504
Table 2: Descriptive statistics for Temperature
Temp
Mean 0.50
Standard Error 0.01
Median 0.50
Mode 0.27
Standard Deviation 0.18
Sample Variance 0.03
Kurtosis -1.12
Skewness -0.05
Range 0.80
Minimum 0.06
Maximum 0.86
Sum
362.1
3
Count 731
Table 3: Descriptive statistics for Humidity
Hum
Mean 0.63
Standard Error 0.01
Median 0.63
Mode 0.61
Standard Deviation 0.14
Sample Variance 0.02

8
IMPACT OF ENVIRONMENTAL FACTOR ON BIKE SHARING PROGRAM IN
WASHINGTON, D.C.
Kurtosis -0.06
Skewness -0.07
Range 0.97
Minimum 0.00
Maximum 0.97
Sum
458.9
9
Count 731
Table 4: Descriptive statistics for Wind speed
Windspeed
Mean 0.19
Standard Error 0.00
Median 0.18
Mode 0.11
Standard Deviation 0.08
Sample Variance 0.01
Kurtosis 0.41
Skewness 0.68
Range 0.49
Minimum 0.02
Maximum 0.51
Sum
139.2
5
Count 731
Table 5: Descriptive statistics for Count of bike rental
Cnt
Mean 4504.35
Standard Error 71.65
Median 4548.00
Mode 1162.00
Standard
Deviation 1937.21
Sample Variance
3752788.2
1

9
IMPACT OF ENVIRONMENTAL FACTOR ON BIKE SHARING PROGRAM IN
WASHINGTON, D.C.
Kurtosis -0.81
Skewness -0.05
Range 8692
Minimum 22
Maximum 8714
Sum 3292679
Count 731
Quality of sample data
Use of good quality of sample data is the primary requirement for attaining a robust
result. Purpose of the statistical analysis is to evaluate impact of environmental factor on
behavior of bike rental. For this, sample data related to bike rental are collected for the two
corresponding years 2011 and 2012 from the system of Capital Bikeshare in Washington D.C.
The additional information related weather and seasonal information are extracted from relevant
websites and then added to the rental data set. For any data set, the researcher first needs an
overall summary measure of the related variable. Important summary statistics related to any
variable are arithmetic mean, median, standard deviation, coefficient of variation and others
(Holcomb 2016). Descriptive statistics for the variables temperature, humidity, wind speed and
count of bike rental are presented in Table 2, Table 3, Table 4 and Table 5. Using the mean and
standard deviation, critical z value and probability density function for each of the numerical
variables are estimated. The critical z values and corresponding PDF are then used to check for
normality of the distribution.

10
IMPACT OF ENVIRONMENTAL FACTOR ON BIKE SHARING PROGRAM IN
WASHINGTON, D.C.
-3 -2 -1 0 1 2 3
0
0.5
1
1.5
2
2.5
Temparature
Z value
PDF
Figure 1: Normal distribution for Temperature
-5 -4 -3 -2 -1 0 1 2 3
0
0.5
1
1.5
2
2.5
3
Humidity
Z value
PDF
Figure 2: Normal distribution for Humidity

11
IMPACT OF ENVIRONMENTAL FACTOR ON BIKE SHARING PROGRAM IN
WASHINGTON, D.C.
-3 -2 -1 0 1 2 3 4 5
0
1
2
3
4
5
6
Windspeed
Z value
PDF
Figure 3: Normal distribution for Wind Speed
-3 -2 -1 0 1 2 3
0
0.00005
0.0001
0.00015
0.0002
0.00025
Count of Bike Rental
Z value
PDF
Figure 4: Normal Distribution for Count of Rental Bikes
Standard statistical theory states that a sample data is considered to be normally
distributed if most of the values in the data series are concentrated close to the vertical axis
meaning the bell-shaped distribution curve has a relatively smaller width (Alizadeh Noughabi
2017). For the sample data representing temperature, humidity, wind speed and count of bike