1

Electric Vehicle Market vs Gas Engine Vehicle Market

This study will focus on the rapidly growing industry for electric vehicles. Although
electric vehicles have been around for way longer than most people know, the industry has only
really just started to take off. Sales of EVs have grown in the last decade and have since boomed
in recent years, with more and more consumers switching over from traditional gas powered cars
to fully electric vehicles. The reason for this is the hope for a decline in the widespread
dependency on a fuel that causes a large amount of carbon emissions, which has become a
problem in today’s world. Following an industry analysis that reviews the current state of gas
and electric vehicles, this study will dive into how demographic factors play into the demand for
this growing industry, and how the problems electric cars face can be tackled.

Gas Vehicles
The Gasoline Vehicle market began to introduce itself in the United States in 1892. This
change and evolution from using horses as transportation to now driving a gas powered vehicle
started a trend that is still around to this day. This movement solved many problems for citizens
during that time. For one, it ultimately gave people more personal freedom and allowed access to
more jobs and services. As more and more gas powered vehicles started to hit the roads, better
transportation conditions, i.e. roads and highways, began growing and expanding throughout the
country.
The Gas Vehicle market was such a hit that in 1920, 9 million vehicles powered by
gasoline were on the road. In 2020, the global automotive natural gas vehicle market demand
was valued at 29,793.76 thousand units and is expected to grow at a compound annual growth
rate of 3.3% from 2021 to 2028. The table below shows the U.S. motor gasoline consumption by
the transportation sector from 2008 to 2020.

Table 1

Year Motor Gasoline Consumption

(in 1,000 barrels per day)

2008 8,834.37

2012 8,525.18

2016 8,972.95

2020 7,736.26
Statista - US Transportation Sector Gasoline Consumption
 2

There are many clear trends in this data, such as the increase in Motor Gasoline
Consumption from 2008 to 2016. This shows that during that time, gasoline power automobiles
were still on a steady increase, burning more and more fuel. From 2020 to 2021, Electric
Vehicles began to boom with sales increasing by 85%, thus, apart from other reasons, resulted in
a decrease of the motor gasoline consumption. In that same time period, plug-in electric vehicles
(PHEVs) sales nearly doubled with an increase in sales of 138% over the previous year.

Electric Vehicle Industry Background

In 2020, electric/hybrid vehicle supply chains were disturbed by COVID-19. This
delayed the production of products in the industry. This, coupled with a high unemployment rate,
minimized the group of potential customers. Eventually, as oil prices skyrocketed and the
economy recovered, demand for electric vehicles increased. Growing concerns for the
environmental impact of carbon emissions also boosted the industry’s market share in car
manufacturing.

Industry revenue is expected to increase as technology advances and long-term

environmental concerns grow. In addition, battery prices are expected to fall and industry
products are predicted to decrease as competition grows in the category. Industry revenue has
grown at an annualized rate of 19.1% to $26.6 billion in the last 5 years. In 2022 alone, an
increase in revenue of 25.3% has been observed as the economy recovers post COVID-19.

Tax incentives, technological advancements, and government assistance are expected to

expand the industry’s revenue over the next five years. This revenue increase is predicted to be
annualized at a rate of 17% leading to $58.3 billion over the years to 2027.

Graph 1

From IBISWorld Report on Hybrid & Electric Vehicle Manufacturing

 3

Table 2

Brand Total Vehicle Sales EV Sales Brand % EV Sales

Tesla 197,517 197,517 100.00%

Smart 1,276 1,219 95.53%

Fiat 15,521 2,250 14.50%

BMW 311,014 22,926 7.37%

Porsche 59,116 3,058 5.17%

Chrysler 165,964 7,062 4.26%

Volvo 98,786 4,091 4.14%

Subaru 680,135 0 0.00%

Evadoption - US EV Sales Percentages of Total Vehicle Sales by Brand

The chart above shows the 2018 EV sales as a percent of each auto brand’s total vehicle
sales in the US. Tesla, the largest EV company by market cap ($749.15 B) is currently the only
brand that has 100% EV Vehicle Sales. Falling shortly after is a company many wouldn’t expect,
Smart, with 95.53% EV Vehicle Sales. As the data shows, more and more vehicle companies are
moving into the Electric Vehicle Market. When comparing this data to data from ten years ago,
these numbers would not be anywhere close to what they are now. The EV Market is one that
has taken off over the past years and does not seem to be slowing down anytime soon.

Graph 2
 4

The graph on the previous page highlights the increase in public charging locations/ports
as of 2020, showing how accessible finding a place to charge is in the US. Currently, there is a
much larger share of Level 2 chargers available compared to Level 1 and DC Fast Charging
ports. Overall, these fast charging ports are better for consumers due to the time saved by using
them. DCFC’s can charge cars in as little as 20-30 minutes, while traditional Level 1 and 2
chargers take much longer. Most consumers who use those chargers have to wait an extended
period of time, some overnight.

Statement of the Business Problem

The problem facing the hybrid and electric vehicle industry today is an amalgamation of
the practicality issues that come with owning an electric vehicle as opposed to a gas powered car.
The issues that consumers of electric vehicles face include: the range one can travel on a single
charge being too low, it can take an unmanageably long time to charge an EV, there are not
always chargers available when they are needed, the price of EV’s largely remains too high for
lower income drivers, and finally managing the power required for people to consistently charge
their cars.
According to a Bloomberg article by Kyle Shock from September 2022, electric vehicles
are consistently selling at a significantly higher price than originally advertised due to dealer
adjustments for supply and demand. “The pragmatic Chevrolet Bolt — purportedly the most
affordable EV in the US at $26,595 — was selling for almost one-third more this summer, at
$34,874 on average” (Shock pgph. 5) While we are not in danger of crashing any power grids in
the United States through charging of EV’s right now, if we continue to see this rapidly
accelerating growth in demand for electric cars and do not adapt our energy allocation practices,
there could be problems in the future.
The purpose of this study is to identify the areas of the EV market that could cause the
success of a transition to electric transportation from gas powered vehicles to falter along the
way, understand what is causing these problems, and seek solutions to the problems in this
market.

Sample
The data used for this study was collected using MTurk. We set MTurk parameters so
that respondents were only drawn from the United States. Once our data was combed through
and cleaned, there were 179 complete and reliable responses. Out of these respondents 63% of
them described their living arrangement as urban, 19.3% as suburban, and 17.5% as rural. The
distribution is shown on the next page:
 5

Graph 3

Table 3

Religion Education Ethnicity Politics

Protestant 8 Less than 0 American 7 Republican 83

High School Indian/Alaska
Native

Catholic 143 High School 8 Asian 6 Democrat 84

Jewish 6 Some College 5 Black/African 4 Independent 5

American

Muslim 7 Associate 4 Latino 2 Other 1

Degree

Other 4 Bachelor’s 129 Native

Degree Hawaiian/ 1
Pacific islander

None 3 Professional 26 White 152

Degree

Doctorate 1
 6

Our sample mostly consists of predominantly white catholic men, who are split between
their political affiliation. In terms of education, our sample consists of very educated individuals,
most having a Bachelor’s Degree. This value is exponentially higher than the United States
average of 34.9%. After looking at the demographics, it is easy to say that our sample is diverse
and will ultimately provide great data. We will be able to examine the attitudes people of many
different backgrounds have towards electric vehicles.

Hypotheses and Results

We ran a handful of statistical tests to identify which factors most impact the opinions
towards fully electric vehicles. The two categories of tests we ran on this data include a priori
hypotheses and post hoc hypotheses. A priori hypotheses are less likely to happen by chance than
post hoc hypotheses because they are based on prior knowledge and reasoning, whereas post hoc
hypotheses are based on observations after the fact. There is a danger that post hoc hypotheses
may happen by chance, as they are often formulated to fit the observed data rather than being
based on pre-existing knowledge or theories. Therefore, those reading this report should keep in
mind that results could be less accurate if they are post hoc hypothesis tests.
The statistical tests run in this report are in the order as follows: independent samples and
paired samples t –tests, the 2 independent samples (Mann Whitney U) and 2 related samples
(Wilcoxon W) tests, Pearson and Spearman correlations, CHI Square Crosstabs, CHAID, and
regression. These tests include a combination of independent and dependent variables, each
aimed at uncovering different results.

Independent Samples T-Test

A Priori Hypothesis: We wanted to use location as a nominal variable so we could
determine if marketing to certain regions would be more effective. The interval variable we
chose was an opinion on whether or not respondents think electric or gas vehicles perform better.
We assumed that those who lived in a certain location would share common opinions on
gas/electric vehicle performance. This is because the terrain of certain regions might affect
opinions on which car would perform best. We anticipated that those in urban settings would
believe that electric cars perform better.

H1: Opinions on what car type is superior regarding performance differs on location.

The dependent variable was measured on a 5-point scale with (1) being that gas powered
vehicles are superior and (8) being that electric vehicles are superior. After running the test, we
can conclude that we should retain the null hypothesis, meaning that when people choose which
car type is superior, there is no difference between performance and location. We can conclude
this because our 2-sided sample test emerged with a p-value of .106 (table 4). We compared this
value with a significant alpha of .05. Because it is greater, we retain the null hypothesis.
 7

Table 4

Post Hoc Exploratory Hypothesis: We ran nine post hoc tests to determine if responses
differed by location when it comes to believing that gas or electric vehicles were superior in
certain areas. The independent variable was the nominal variable, location, and the dependent
variables were factors that respondents believed gas or electric cars were superior in. The graphic
below shows the two tests that showed a difference in response based on location.

Table 5
 8

Graph 4

At a significance level of .05, two tests resulted in the rejection of the null hypothesis.
The first test was run to determine if responses differed by location on their opinion of
gas/electric vehicle dealership quality. Respondents that lived in rural areas thought that electric
vehicles were superior in how their dealerships were run. This was determined by observing the
p-value of .042 in table 5 and the results in table 6. The other test that resulted in the rejection of
the null was seeing if location affected whether or not respondents thought gas/electric vehicles
were superior in terms of environmental impact. At a p-value of .014, respondents in
urban/suburban areas believed that electric vehicles were superior when it comes to
environmental impact. This is shown in table 5 above and 7 below.

Graph 5
 9

Paired Samples T-Test

A Priori Hypothesis: As a marketer, it is extremely important to know which attributes
are important to your customer when working on R&D projects for your product/brand. Paired
samples t-test is a useful tool for learning about the differences between the means of two related
groups and for making inferences about the populations from which the groups were sampled.
We developed a broad hypothesis to see if there is a correlation between two aspects regarding
vehicles.

H2: Environmental Impact and Performance differ in importance.

H3: Environmental Impact and Range differ in importance.

Table 6
(1 Not at all Important , 1st 2nd
5 Extremely Important)

Variable T-Value P-Value Mean Mean

Environmental Impact / Performance -1.841 .067 3.84 3.98

Environmental Impact / Range -1.550 .123 3.84 3.96

Both hypotheses are not supported through this test, meaning that environmental impact
and performance don’t differ in importance as well as environmental impact and range. We can
conclude this because each p-value is greater than the alpha of .05. Although they weren’t
supported, we are still able to conclude that consumers (at least in our sample) feel that
environmental impact, performance, and range are all important factors when considering
vehicles, whether it be electric or gas. Based on the mean values, the respondents of our survey
have the highest attitude toward performance, meaning that they feel it is extremely important.
Marketers can learn from this and produce faster and more powerful electric vehicles. This may
increase traffic in their stores as well as increase sales.
Another piece of information we can learn from this test is that when producing a vehicle
(electric or gas), focus a lot of your attention on the range capabilities. This aspect’s level of
importance was just slightly below that of performance’s.
Post Hoc Exploratory Hypothesis: We ran a post hoc exploratory test in which we
compared the importance of every vehicle aspect included in the study with all other aspects. In
doing so, we created five new pairs to see if their importance correlates or differs from each
other. Out of the five tests ran, three of them showed to be significant. This means that when
paired together, three of the variables had significant differences. The following data shows the
relationships.
 10

Table 7
(1 Not at all Important , 1st 2nd
5 Extremely Important)

Variable T-Value P-Value Mean Mean

Delivery Time / Price -3.620 <.001 3.73 4.06

Brand / Aesthetic Appeal .316 .025 3.98 3.81

Safety / Environmental Impact .363 .016 4.03 3.83

Range / Price .062 .237 3.96 4.05

Dealership / Delivery Time .235 .518 3.79 3.73

As you can see, of the five paired samples, three are exhibiting significance due to their
low p-value. Of those three, Delivery Time and Price prove to be the most significant. Delivery
Time had a mean of 3.73 and Price had a mean of 4.06, meaning that of our sample, respondents
believe that the price of the vehicle is more important than the delivery time. When buying a
vehicle, customers are more concerned about the overall price of the purchase rather than the
time it takes to get delivered to their house. In this case, Tesla’s marketing strategy aligns
perfectly in order to attract the right customers. Tesla markets their vehicles online via their
websites. It Is important to know that customers would rather wait longer to receive their vehicle
and pay a lower price than receive it quicker and pay more.
The next most significant pair is Safety and Environmental Impact with a p-value of .016.
The mean values in this case were as follows; Safety 4.03, Environmental Impact 3.83.
Customers are more concerned with the safety of their vehicle than the impact it has on the
environment. This data wasn’t extremely shocking to us as most Americans would prefer to have
a safer car.
The third and final significant pair was Brand and Aesthetic Appeal. In this pairing,
Brand had a mean of 3.98 and Aesthetic Appeal of 3.81. Though not extremely different in
value, our sample believes that brand is more important than the aesthetic appeal of the vehicle.
Though these two variables go hand in hand in the car industry, the relationship between the two
is interesting to see.

2 Independent Samples (Mann-Whitney U)

 11

A Priori Hypothesis: A nominal variable used in our survey helped to eliminate the
factor of price in people’s decision to purchase an electric vehicle or a gas-powered vehicle. We
made the decision to test this variable against people's political ideology because this can be an
important trait by which marketers develop market segments, especially in large industries (like
the automobile industry) that have a great effect on major environmental, societal, and economic
factors. We are interested in finding if there is a significant difference in the overall attitude
towards owning an electric between people who are more liberal and those who are more
conservative. A 2 independent samples test between the nominal variable which asks: “If the
purchasing price for the average electric vehicle was exactly the same as the average gas
powered vehicle, would you choose an electric vehicle?” and the interval variable that classifies
political ideology as 1 being very liberal and 7 being very conservative will allow us to make a
conclusion about this hypothesis.

H4: People’s decision to purchase an electric vehicle if it were priced equally to a gas-
powered vehicle in their price range differs based on political ideology.

The test yielded a p-value of .361, and tested against an alpha value of .05, this
hypothesis is clearly not supported by the data. 35.2% of our respondents classified themselves
as conservative, and while we expected there to be a significant portion of these people to remain
loyal to gas-powered vehicles, the vast majority of them along with those from all other political
ideologies answered yes, they would purchase an electric vehicle if they were priced equally to
gas-powered vehicles.
Post Hoc Exploratory Hypothesis: We ran 9 post hoc tests where the independent variable was
the nominal gas powered/electric preference and the dependent variables were interval or ratio. 5
of these tests were run based on responses to the personality type questions.

Table 8
Variable Yes No
MW-U P-Value Mean Mean

I see myself as someone who is sympathetic, warm. 1,358 0.629 85.88 92.53

I see myself as someone who is disorganized, careless. 1,897 0.027 89.4 63.16

I see myself as someone who is extraverted, enthusiastic. 1,673.5 0.272 87.94 74.92

I see myself as someone who is anxious, easily upset. 1,707 0.206 88.16 73.16

I see myself as someone who is conventional, uncreative. 1,573.5 0.313 87.28 75.08
 12

The data shows that the relationship between the decision to purchase an electric vehicle
over a gas vehicle at the same price and the degree to which people see themselves as
disorganized and careless is significant with a p-value of 0.027. The mean response for “I see
myself as someone who is disorganized, careless” was 5.1 showing that people who generally
see themselves as disorganized and careless will more often decide to buy an electric car if it
were priced equally to a gas-powered car. Without significant data regarding the other four
personality types, it is hard for us to make a marketing recommendation to target one group over
the others. However, we can be confident that people who are more apathetic in general will
likely purchase an electric car if the price is able to match the car that they may currently have.

2 Related Samples (Wilcoxon Signed Ranks)

A Priori Hypothesis. A critically important question for electric car producers is how
popular different brands of cars are. Presumably, companies will gain higher profits if they
produce the brand of car most in demand. We accordingly tested a broad hypothesis.

H5: The various kinds of car brands differ in their degree of popularity.

Here, we did not specify comparisons between only two car brands. We compared every car
brand with every other with the hypothesis that liking would differ. Thus, for the 10 car brands
studied here (Tesla, Rivian, Volvo, Mercedes Benz, Toyota, Chevrolet, Nissan, BMW, Kia, and
Mazda) there were 45 possible comparisons. We conducted all of them, and all but one (the
comparison between Nissan and BMW) yielded significant differences in liking. The tests are
reported in Table 9.

Table 9

Variable P-Value Negative Difference Positive Difference

Tesla - Rivian, 1 Like most, 10 Like least

<.001 163 37
Tesla - Volvo, 1 Like most, 10 Like least
<.001 154 46
Tesla - Mercedes Benz, 1 Like most, 10 Like least
<.001 158 42
Tesla - Toyota, 1 Like most, 10 Like least
<.001 151 49
Tesla - Chevrolet, 1 Like most, 10 Like least
0.000 172 28
Tesla - Nissan, 1 Like most, 10 Like least
0.000 176 24
 13

Tesla - BMW, 1 Like most, 10 Like least 0.000 157 43

Tesla - Kia, 1 Like most, 10 Like least 0.000 182 18

Tesla - Mazda, 1 Like most, 10 Like least 0.000 185 15

Rivian - Volvo, 1 Like most, 10 Like least 0.083 142 58

Rivian - Mercedes Benz, 1 Like most, 10 Like least 0.005 143 57

Rivian - Toyota, 1 Like most, 10 Like least 0.003 124 76

Rivian - Chevrolet, 1 Like most, 10 Like least <.001 153 47

Rivian - Nissan, 1 Like most, 10 Like least <.001 159 41

Rivian - BMW, 1 Like most, 10 Like least 0.000 132 68

Rivian - Kia, 1 Like most, 10 Like least 0.000 167 33

Rivian - Mazda, 1 Like most, 10 Like least 0.000 177 23

Volvo - Mercedes Benz, 1 Like most, 10 Like least 0.269 138 62

Volvo - Toyota, 1 Like most, 10 Like least 0.198 128 72

Volvo - Chevrolet, 1 Like most, 10 Like least <.001 147 53

Volvo - Nissan, 1 Like most, 10 Like least <.001 154 46

Volvo - BMW, 1 Like most, 10 Like least <.001 122 78

Volvo - Kia, 1 Like most, 10 Like least 0.000 158 42

Volvo - Mazda, 1 Like most, 10 Like least 0.000 178 22

Mercedes Benz - Toyota, 1 Like most, 10 Like least 0.856 128 72

Mercedes Benz - Chevrolet, 1 Like most, 10 Like least <.001 154 46

Mercedes Benz - Nissan, 1 Like most, 10 Like least <.001 150 50

Mercedes Benz - BMW, 1 Like most, 10 Like least <.001 123 77

Mercedes Benz - Kia, 1 Like most, 10 Like least 0.000 165 35

Mercedes Benz - Mazda, 1 Like most, 10 Like least 0.000 181 19

Toyota - Chevrolet, 1 Like most, 10 Like least <.001 156 44

 14

Toyota - Nissan, 1 Like most, 10 Like least <.001 164 36

Toyota - BMW, 1 Like most, 10 Like least <.001 125 75

Toyota - Kia, 1 Like most, 10 Like least 0.000 160 40

Toyota - Mazda, 1 Like most, 10 Like least 0.000 177 23

Chevrolet - Nissan, 1 Like most, 10 Like least 0.574 153 47

Chevrolet - BMW, 1 Like most, 10 Like least 0.006 118 82

Chevrolet - Kia, 1 Like most, 10 Like least <.001 148 52

Chevrolet - Mazda, 1 Like most, 10 Like least 0.000 169 31

Nissan - BMW, 1 Like most, 10 Like least 0.027 114 86

Nissan - Kia, 1 Like most, 10 Like least <.001 149 51

Nissan - Mazda, 1 Like most, 10 Like least 0.000 165 35

BMW - Kia, 1 Like most, 10 Like least 0.001 159 41

BMW - Mazda, 1 Like most, 10 Like least <.001 172 28

Kia - Mazda, 1 Like most, 10 Like least <.001 165 35

This table reports the results of the 2 Related Samples Wilcoxon Signed Ranks Test, a
within subjects test for ordinal data. In this test, subjects rank their liking for car brand. Then the
rank of the second car brand is subtracted from the rank of the first. So in the first table example,
the formula was Tesla Rank - Rivian Rank. When that calculation was done, the difference
between the two was negative 163 times and positive 37 times. So Rivian had the higher rank for
163 subjects and Tesla had the higher rank for 37 subjects. But notice that higher numbers mean
less liking. Rivian has the higher number (was liked less than Tesla) 163 times. Only 37 subjects
liked Rivian more than Tesla.
The car brands have been entered in the table in the order of their popularity. Tesla was
more popular than all other meats. Rivian was more popular than all other brands except Tesla.
Volvo was more popular than than all brands other than Rivian and Tesla. Mercedes Benz was
more popular than Toyota, Chevrolet, Nissan, BMW, Kia and Mazda. Toyota was more popular
than Chevrolet, Nissan, BMW, Kia and Mazda. Nissan was less popular than Chevrolet, but
more popular than BMW Kia and Mazda. BMW was more popular than Kia and Mazda, and Kia
was more popular than Mazda. Finally, the least popular brand was Mazda. These results show
quite clearly that the electric car industry needs to focus on the development of Tesla products,
then electric Rivian, then Volvo, then Mercedes Benz, then Toyota, then Chevrolet, then Nissan,
then BMW, then Kia, then finally Mazda.
 15

Pearson Correlation
A Priori Hypothesis: Based on political stereotypes, we thought that liberal-leaning
individuals would favor a full transition to electric vehicles. Therefore, we ran a pearson
correlation on two interval types of data to test this.

H6: The people who consider themselves ‘Liberal’ are related to people who agree that
electric vehicles should be fully transitioned to

The data came back in favor of conservative-leaning individuals believing that we should
fully transition to electric vehicles. This was proven with a p-value of .035 in comparison to an
alpha of .05. Furthermore, an average of 1.84 in political ideology favors conservatism since our
scale for political ideology is from 1-7. A one on the political ideology scale shows heavy
liberalism and a 5 on the scale shows heavy conservatism. Lastly, an average of 4.93 for
believing that we should transition to electric vehicles is high. This is because that variable is on
a 1-5 scale where 5 favors a heavy belief in transitioning.

Again, based on political stereotypes, we wanted to see if liberal-leaning individuals

favored the ownership of electric vehicles. We ran another pearson correlation on our data to
determine this.

H7: The people who consider themselves ‘Liberal’ are related to the people who own an
electric vehicle.

The data came back in favor of conservative-leaning individuals owning electric cars. At
a p-value 0f .002, a strong correlation was established. The average mean for political ideology
was 4.85. Since our scale for political ideology was from 1-7, the average of 4.85 means that the
data represents a majority of conservative individuals. This, paired with the mean for answering
yes to owning an electric car being .77 further establishes a correlation between political
ideology and owning an electric car (Table 9)

Table 10

Variable P-Value Mean 1 Mean 2

Liberal/Transition to Electric Vehicles .035 1.84 4.93

Liberal/Owning Electric Vehicle .002 .77 4.85

 16

Spearman Correlation
A Priori Hypothesis: As discussed in prior tests, politics is quite closely related to the
electric vehicle industry and thus can have major marketing implications. We previously tested
one’s political ideology – i.e. liberal to conservative – against people’s opinion on whether or not
we should make a full transition to electric vehicles. The results of this were somewhat
surprising in that across all ideologies, people almost entirely answered yes to this question. So,
in this test, we will test the variable that asked people their degree of political involvement with
the same question about transitioning to electric vehicles completely with a Spearman
Correlation test to see if the two are related.

H8: There is a correlation between political involvement and the degree to which one
believes that we should fully transition to electric cars.

The Spearman correlation test came back with a p-value less than .001, which proves that
the correlation coefficient of .468 is significant. The variable for political involvement is on a
seven point scale, 1 being “Very Uninvolved Politically” and 7 being “Very Involved
Politically”, and the variable testing people’s opinion on if we should transition fully to electric
vehicles is on a five point scale, 1 being “strongly disagree” and 5 being “strongly agree”. This
means that the strong positive correlation suggests that the more politically involved a person is,
the stronger their belief in the transition to a world without gas powered vehicles. This is a
logical conclusion as the effect of fossil fuels on climate change and the fact that they are rapidly
becoming more scarce is a hot-button topic in the political landscape today. Those that are
consuming political news often and forming their own opinion based on happenings in the
political landscape will be very familiar with the need for a shift from fossil fuel powered
vehicles, as well as players in the electric vehicle market like Elon Musk of Tesla who have
become very politically polarizing figures.

CHI Square Test

A Priori Hypothesis: Marital status can reveal a great deal about consumers’ buying
habits. We want to discover if there is a relationship between one’s marital status, i.e.
1=Divorced, 2=In a domestic partnership, 3=Married, 4=Single, 5=Widowed, and whether or not
they currently own an electric vehicle . We will run a CHI Square Crosstabs test in order to
examine these two nominal variables. If the test reveals that there is a significant correlation
between these two variables, we can make conclusions about the family dynamics and living
situations of people who do or do not own electric cars.

H9: Owning an electric car is related to marital status

 17

The Pearson Chi-Square value from this test was 19.958 with a p-value of less than .001.
This tells us that when the null hypothesis (owning an electric car is not related to marital status)
is assumed true, we will get a chi square value of 19.958 or greater less than one in one thousand
times. This can be interpreted to mean that there is significant difference between the observed
values and expected values for respondents’ of each marital status to the question: “Do you own
an electric car?”

Table 11
Do you own an
electric vehicle?

Marital Status Count No Yes

Divorced Observed 2 0
Expected .5 1.5

Domestic Partnership Observed 2 0

Expected .5 1.5

Married Observed 26 114

Expected 32.5 107.5

Single Observed 8 15
Expected 5.3 17.7

Widowed Observed 1 0
Expected .2 .8

The greatest difference between observed and expected counts in this Chi-Square test
were for those that are married. These particular respondents answered that they do own an
electric car 114 times, when it was expected that there would only be 107.5 married people
responding that they have electric cars. Simply comparing those that are married versus those
that are single can help to reveal interesting insights as well. 81.4% of married people own
electric cars, while only 65.2% of single respondents do. This can be useful to marketers as they
can begin to target people who are married more heavily, perhaps by using seasonal
advertisements which pose electric cars as excellent Christmas, Valentines day, or anniversary
gifts. Marketers of electric vehicles can also begin to position their cars as excellent family
vehicles.
 18

CHAID Analysis
An important factor for seeing problems in the electric car industry is whether or not a
test subject owns a car. This group has a better knowledge about everything related to cars, and
can give better ideas to marketers concerning the EV industry compared to people who don't own
cars. For this CHAID analysis, many variables tested for importance came back with no valid
cases. This led us to decide to change our level of significance from .05 to .1, in order to capture
more of the market. We also decided to set our minimum number of subjects per child cell at 5
due to the small size of our sample.

Tree 1

After running the analysis, the most important predictor in our data set was household
income. We found that out of the car-owning households that make $45,000-59,000 per year,
100% of them (14.5% of the population) would want their EV to charge in less than 6 hours.
Only 63.6% (6.4% of the population) of car-owning households would want a greater than 6
hour charging time. We can draw from this that low to mid-range income households would
prefer an EV (if they owned one) to have a relatively quick charging time.
 19

In the world of electric cars, There are currently different ways for charging. There are
level 1 chargers, which take approximately 11-20 hours to get a full charge. These are the
chargers that EV manufacturers supply to their customers. There are also level 2 chargers, which
take about 3-8 hours for a full charge. These chargers have to be purchased for a premium price,
which is most likely going to be out of the range of a low to mid income household. There also
exists level 3, which is the most costly but most effective way of charging.
We can conclude from this CHAID analysis that there exists a problem in the EV
industry that can be supported through data- chargers need to be lowered in price in order to
attract more lower-middle class EV car consumers. Right now, the economic state of these
potential EV consumers is limiting them from making a purchase. If Level 2 chargers were
cheaper, there is a likelier chance that lower to middle income households would purchase an
electric car.

Regression
We decided to run a regression in relation to our CHAID analysis. From the CHAID, we
can create consumer personas of people that could purchase an EV. For example, the lower-
middle class household would be more willing to pay for an electric car if the charging option
was cheaper than it currently is for a 3-8 hour charge. But, how much would they be willing to
pay?
In our regression, the dependent variable was the amount of money consumers were
willing to pay on an electric vehicle. Our independent variables were total household income and
preferred maximum time spent charging an EV to 100%. We used the enter method for
conducting our regression.
Table 12

From the results on the previous page, we saw that as the amount of time spent charging
an EV went up by one hour, the price of the car increases by $4,433.12 for consumers. Also, for
 20

every $10,000 increase in household income, a consumer would be willing to pay $314.29 more
for the electric car.
We then took the range of income and charging time indicated in the CHAID analysis for
our low-middle class household that would want their car charged in less than 6 hours, and found
the price range that this segment would be willing to pay for an electric car.

Table 13: Bottom of Range- 1 hour to charge and $45,000 household income

Beta x xX

13672.046 13672.046 Coefficient

4433.118 1 4433.118 Charging time (hrs)

314.290 45 14143.05 Income (000's)

$ 32,248.21 Price

Table 14: Top of Range- 6 hours to charge and $59,000 household income

Beta x xX

13672.046 13672.046 Coefficient

4433.118 6 26598.708 Charging time (hrs)

314.290 59 18543.11 Income (000's)

$ 58,813.86 Price

We can see that any consumer that falls into this low-middle class range would pay
anywhere from $32,248.21 to $58,813.86, highlighting a recurring problem: EVs are just too
expensive for the majority of the US population (being low-middle income households) with
current market conditions. According to Kelley Blue Book, the average electric car costs
$62,876, being over $4,000 of the highest amount this low-middle income household segment is
willing to pay.
 21

Multidimensional Scaling
Unfortunately, the data collected through our survey did not support the formulation of a
multidimensional scaling plot.

Summary and Conclusion

Through the many tests conducted, we are able to conclude that the Electric Vehicle
Industry is one that is rapidly growing and favored in the United States. There are several
reasons why this market is growing so rapidly, one being that electric cars have become more
affordable in recent years due to advances in technologies. This has made them more accessible
to a wider range of customers.
As noted in Graph 3 as well as Table 3, you are able to see our very diversified sample of
179 respondents. Because our sample was so diversified, we were able to run specific tests to
target certain demographics and learn about their attitudes toward the EV market overall. For
example, in Table 10 overviewing a CHI Square Test, you are able to look at which group of
individuals owns an EV and which does not. From this test, we were able to learn that 81.4% of
married people own electric cars, while only 65.2% of single respondents do. As learned in
Market Analytics, focusing your advertisements to the right market is crucial when not only
trying to grow a business, but when improving your business. This ultimately starts with
increasing your response rates. EV marketers can use this information to target the right group,
saving them money and earning them market share.
When discussing the EV Industry as a whole, we are able to use the Pearson Correlation
to analyze some data. In just this test, we are able to conclude that American’s favor wanting to
transition to electric vehicles with a mean of 4.93 (scaled 1-5, 5favors a heavy belief in
transitioning). This data alone tells that the vehicle market, at least in the United States, is
starting to greatly transition to all electric vehicles.
The CHAID test that we ran showed that the strongest predictor in the dataset was
household income. We used this variable and the subsequent branch that asks respondents the
maximum full-charge time they would accept for an electric vehicle to develop a regression
analysis that will predict what a person would be willing to pay for an electric vehicle.
 22

Appendix: Study Survey

Electric Vehicle Industry Problems

Start of Block: Informed Consent

InformedConsent
Identification of Investigators & Purpose of Study
You are being asked to participate in a research study conducted by Patrick Hinkle, Jonathan
Patterson, Dominic Castellano, Tyler Girard, and Robert Swain from James Madison University.
The purpose of this study is to learn about how various demographic and psychographic attributes
affect consumer preferences. This study will contribute to the researchers’ completion of a data
analysis class project.

Research Procedures
This study consists of an online survey that will be administered to individual participants through
MTurk and Qualtrics (an online survey tool). You will be asked to provide answers to a series of
questions related to the electric vehicle industry.

Time Required
Participation in this study will require about 15 minutes of your time.

Risks
The investigators do not perceive more than minimal risks from your involvement in this study
(that is, no risks beyond the risks associated with everyday life).

Benefits
Potential benefits from participation in this study include compensation paid through MTurk.

Confidentiality
The results of this research will be presented in a written report that meets a class requirement.
While individual responses are anonymously obtained and recorded online through MTurk and
Qualtrics, data is kept in the strictest confidence. No identifiable information will be collected from
the participant and no identifiable responses will be presented in the final form of this study. All
data will be stored in a secure location only accessible to the researchers. The researchers retain the
right to use and publish non-identifiable data. At the end of the study, all records will be destroyed.
Final aggregate results will be made available to participants upon request.
 23

Participation & Withdrawal

Your participation is entirely voluntary. You are free to choose not to participate. Should you
choose to participate, you can withdraw at any time without consequences of any kind. However,
once your responses have been submitted and anonymously recorded you will not be able to
withdraw from the study.

Questions about the Study

If you have questions or concerns during the time of your participation in this study, or after its
completion or you would like to receive a copy of the final aggregate results of this study, please
contact:

Researcher's Name: Jonathan Patterson

Department of Marketing
James Madison University
Email Address: [email protected]

Advisor’s Name: Val Larsen

Department of Marketing
James Madison University
Email Address: [email protected]
Telephone: (540) 568-3858

Questions about Your Rights as a Research Subject

Dr. Lindsey Harvell-Bowman
Chair, Institutional Review Board
James Madison University
Email Address: [email protected]
Telephone: (540) 568-2611

Giving of Consent
I have been given the opportunity to ask questions about this study. I have read this consent and I
understand what is being requested of me as a participant in this study. I certify that I am at least
18 years of age. By clicking on the link below, and completing and submitting this anonymous
survey, I am consenting to participate in this research.

This study has been approved by the IRB, protocol # 23-3825.

 24

o I consent, begin the study (1)

o I do not consent, I do not wish to participate (2)
End of Block: Informed Consent

Start of Block: SURVEY INSTRUCTION

Sex What is your sex?

o Male (1)
o Female (2)
o Other (Please Specify) (3) __________________________________________________

Ethnicity What is your ethnicity?

o American Indian or Alaska Native (1)

o Asian (2)
o Black or African American (3)
o Latino or Hispanic (4)
o Native Hawaiian or Pacific Islander (5)
o White (6)
o Other (please specify) (7) __________________________________________________
 25

Age Age (in years)

18 26 34 43 51 59 67 75 84 92 10
0

Click to write Choice 1 ()

Marriage Marital status

o Divorced (1)
o In a domestic partnership (2)
o Married (3)
o Separated (4)
o Single (5)
o Widowed (6)
o Other (please specify) (7) __________________________________________________

Income Total household income (in thousands of dollars) up to $175,000+

0 18 35 53 70 88 10 12 14 15 17
5 3 0 8 5

Income ()
 26

Household# How many people are there in your household?

0 1 2 3 4 6 7 8 9 10 11

Household ()

Children How many children do you have? (If 0 click the slider at 0.)
0 1 2 3 4 6 7 8 9 10 11

Children ()

EdDegree What is the highest level of school you have completed or the highest degree you have
received?

o Less than high school (1)

o High school graduate (2)
o Some college (3)
o 2 year degree (4)
o 4 year degree (5)
o Professional degree (6)
 27

o Doctorate (7)

Employment Employment status

o Employed full time (1)

o Employed part time (2)
o Unemployed looking for work (3)
o Unemployed not looking for work (4)
o Retired (5)
o Student (6)
o Disabled (7)

Location The place where I live is best described as ...

o Urban (1)
o Suburban (2)
o Rural (3)
 28

Religion What is your religion?

o Buddhist (1)
o Catholic (2)
o Hindu (3)
o Jew (4)
o Muslim (5)
o None (6)
o Protestant (7)
o Other (please specify) (8) __________________________________________________

Religiosity Degree of religiosity

Very Irreligio Somewhat Neither Somewhat Religious Very
irreligiou us (2) irreligious religious religious (6) religiou
s (1) (3) nor (5) s (7)
irreligiou
s (4)

Rate your
degree of
religiosity
o o o o o o o
(1)
 29

PoliticalParty What is your political affiliation?

o Republican (1)
o Democrat (2)
o Independent (3)
o Other (please specify) (4) __________________________________________________

Political Involvement What is your degree of political involvement

Very Uninvolve Somewha Neither Somewha Involve Very
Uninvolve d (2) t Involved t d (6) Involve
d Uninvolve nor Involved d
Politically d (3) Uninvolve (5) Political
(1) d (4) ly (7)

Political
involveme
nt (1)
o o o o o o o

Ideology Which descriptor best fits your political views?

Very Liber Somewhat Moderat Somewhat Conservativ Very
Liber al (2) Liberal e (4) Conservativ e (6) Conservativ
al (1) (3) e (5) e (7)

Politica
l views
(1)
o o o o o o o
 30

Personality:ACENO I see myself as someone who is ...

Strongly Disagree Somewhat Neither Somewhat Agree Strongl
Disagree (2) Disagree Agree Agree (5) (6) y Agree
(1) (3) nor (7)
Disagree
(4)

...
sympathetic,
warm. (1)
o o o o o o o
...
disorganized,
careless. (2)
o o o o o o o
... is
extraverted,
enthusiastic.
o o o o o o o
(3)

... anxious,
easily upset.
(4)
o o o o o o o
...
conventional,
uncreative.
o o o o o o o
(5)

Attention To accurately assess attitudes, we need accurate responses. To demonstrate you are
paying attention, select the second response from the left, Disagree, on this item regardless of your
actual feelings.
Strongly Disagree Somewhat Neither Somewhat Agree Strongly
Disagree (2) Disagree agree Agree (5) (6) Agree
(1) (3) nor (7)
disagree
(4)
 31

I am a
confident
person.
o o o o o o o
(1)

ShopVenue Where do you most like to shop, online or in a store?

Strongly Prefer Somewhat No Somewhat Prefer Strongly
Prefer Online Prefer Preference Prefer a a Store Prefer a
Online (2) Online (3) for Online Store (5) (6) Store.
(1) or Store (7)
(4)

Shopping
preference
(1)
o o o o o o o

SocialMedia Indicate how often you use the following social media.
Never (1) Less than once a About once a More than once
day (2) day (3) a day (4)

YouTube (1)
o o o o
Facebook (2)
o o o o
Snapchat (3)
o o o o
Instagram (4)
o o o o
 32

Twitter (5)
o o o o

Hobbies By moving the hobbies, rank them from 1 (most enjoyable for you) to 7 (least enjoyable for
you)
______ Gardening (1)
______ Playing video games (2)
______ Watching TV (3)
______ Surfing the internet (4)
______ Reading (5)
______ Traveling (6)
______ Shopping (7)

Google Select response 6, Agree, on this item to show you are paying attention.
Strongly Disagree Somewhat Neither Somewhat Agree Strongly
Disagree (2) Disagree Agree Disagree (6) agree (7)
(1) (3) nor (5)
Disagree
(8)

Google
use. (1)
o o o o o o o

YesNo Which of the following do you regularly use?

Use (1) Don't Use (2)

Dental floss (1)

o o
 33

A daily newspaper (2)

o o
Amazon Prime (3)
o o
Diet soda (4)
o o

End of Block: SURVEY INSTRUCTION

Start of Block: Block 2

Q29 Do you own a car?

o Yes (1)
o No (0)

Q30 Do you own an electric car?

o Yes (1)
o No (0)

Q31 What make of electric vehicle do you own? (Check all that apply)

▢ Tesla (1)

▢ Mazda (2)
 34

▢ Volvo (3)

▢ Mercedes Benz (4)

▢ Toyota (5)

▢ Chevrolet (6)

▢ Nissan (7)

▢ BMW (8)

▢ Hyundai (9)

▢ Kia (10)

▢ Other (11)

Q32 Indicate your feelings about the following issues with electric vehicle ownership:
Strongly Somewhat Neither Somewhat Strongly
disagree (1) disagree (2) agree nor agree (4) agree (5)
disagree (3)

The Range of
Electric
Vehicles is too
o o o o o
Short (1)
 35

People
Experience
Trouble
o o o o o
Finding A
Charger (2)

It Takes Too
Long to
Charge An
o o o o o
Electric
Vehicle (3)

Electric
Vehicles are
not
o o o o o
Sufficiently
Powerful (4)

Q33 You have decided to purchase an electric vehicle. Rank the following brands based from 1
(most likely to purchase) to 10 (least likely to purchase).
______ Tesla (1)
______ Rivian (2)
______ Volvo (3)
______ Mercedes Benz (4)
______ Toyota (5)
______ Chevrolet (6)
______ Nissan (7)
______ BMW (8)
______ Kia (10)
______ Mazda (11)

Q34 Using the slider, indicate how much you would be willing to spend on an electric vehicle.
0 25000 50000 75000 10000 12500 15000
0 0 0

Price in Dollars ($) ()

 36

Q35 If the purchasing price for the average electric vehicle was exactly the same as the average gas
powered vehicle, would you choose an electric vehicle?

o Yes (1)
o No (0)

Q36 If you were purchasing an electric car, what is the maximum full recharge time you would
accept?
0 2 4 6 8 10 12

Hours Spent Charging to a 100% battery ()

Q37 Rate the importance of the following factors when buying a car:

Not at all Slightly Moderately Very Extremely

important important important important important
(1) (2) (3) (4) (5)

Delivery Time
(1)
o o o o o
Price (2)
o o o o o
Brand (3)
o o o o o
Dealership (4)
o o o o o
 37

Fuel cost (5)

o o o o o
Performance
(6) o o o o o
Aesthetic
Appeal (7) o o o o o
Range (8)
o o o o o
Safety (9)
o o o o o
Environmental
impact (10) o o o o o

Q38 For each attribute, indicate the degree to which gas-powered or electric-powered vehicles are
superior:
Gas Powered (-1) (0) (1) Electric
Vehicle Powered
Superior (-2) Vehicle
Superior (2)

Delivery Time
(1) o o o o o
Price (2)
o o o o o
Brand (3)
o o o o o
 38

Dealerships (4)
o o o o o
Fuel Cost (6)
o o o o o
Performance
(7) o o o o o
Aesthetic
Appeal (8) o o o o o
Range (9)
o o o o o
Safety (10)
o o o o o
Environmental
Impact (11) o o o o o

Q39 How much range would an electric vehicle need to have for you to purchase one?
0 100 200 300 400 500 600 700 800

Mile Range Per Charge ()

Q40 I believe we should fully transition to electric vehicles.

Strongly Somewhat Neither Somewhat Strongly
disagree (1) disagree (2) agree nor agree (4) agree (5)
disagree (3)
 39

Should
transition to
electric
o o o o o

End of Block: Block 2

Start of Block: Block 3

Code Here is your MTurk code: ${e://Field/Random%20ID}

Copy this value so you can paste it into MTurk. After you have copied this Mturk code, please click
the >> button to submit your survey.

End of Block: Block 3

Works Cited

Automotive Natural Gas Vehicle Market Size Report, 2021-2028,

https://www.grandviewresearch.com/industry-analysis/automotive-natural-gas-vehicles-
market.

Can the Nation's Electrical Grid Support Electric Cars? https://cars.usnews.com/cars-

trucks/features/can-the-nations-electrical-grid-support-electric-cars.

Carlier, Mathilde. “U.S. Motor Gasoline and Distillate Fuel Consumption 2020.” Statista,
28 Sept. 2022, https://www.statista.com/statistics/189410/us-gasoline-and-diesel-
consumption-for-highway-vehicles-since-1992/.

“EBSCOhost Research Platform: EBSCO.” EBSCO Information Services, Inc. |

Www.ebsco.com, https://www.ebsco.com/products/ebscohost-research-platform.
 40

“Electric Car Industry.” IBIS World, https://my.ibisworld.com/us/en/industry-

specialized/od4516/industry-at-a-glance#executive-summary.

“The History of the Electric Car.” Energy.gov, https://www.energy.gov/articles/history-

electric-car.

“New Plug-in Electric Vehicle Sales in the United States Nearly Doubled from 2020 to
2021.” Energy.gov, https://www.energy.gov/energysaver/articles/new-plug-electric-
vehicle-sales-united-states-nearly-doubled-2020-
2021#:~:text=EV%20sales%20grew%20by%2085,3%25%20during%20the%20same%20p
eriod.

Randall, Tom. “US Electric Car Sales Reach Key Milestone.” Bloomberg.com, Bloomberg,
9 July 2022, https://www.bloomberg.com/news/articles/2022-07-09/us-electric-car-sales-
reach-key-milestone.

“Us Ev Sales Percentages of Total Vehicle Sales by Brand.” EVAdoption, 26 Apr. 2022,
https://evadoption.com/ev-sales/evs-percent-of-vehicle-sales-by-brand/.

user, A. (2022, May 17). What is the average price of an electric car? Jerry. Retrieved
December 16, 2022, from https://getjerry.com/questions/what-is-the-average-price-of-an-
electric-car

rocket55. (2022, October 17). The difference between level 1 & 2 EV Chargers.
EvoCharge. Retrieved December 16, 2022, from https://evocharge.com/resources/the-
difference-between-level-1-2-ev-chargers/