Human Behaviour Statistics with Video Games - Research Paper Example

4. Chapter Four: Findings and Discussion

This chapter contains all the results of the methodology study conducted to find out if the following hypothesis is true:

• H1: The live broadcast investment of a game is positively correlated with the game's sales, which the number of channels that the viewer sees the game is proportional to their purchase intention.
• H2: The number of viewers and the number of real-time comments of a game will positively affect game sales.
• H3a: Compared to non-multiplayer online games, live streaming of multiplayer online games will positively affect customer’s willingness to buy then increase game sales.
• H3b: Compared to other types of games, computer games based on story content will negatively affect customers' willingness to buy and then reduce game sales.
• H4: The audience impression of the streamer will positively affect the willingness to buy and thus increase product sales.

This chapter will also have a discussion section in which the analysis provided will be used to prove the hypothesis mentioned above. It will also employ the use of charts and graphs to complement the discussion presented.

The data was obtained from the questionnaire survey conducted by the researcher, as mentioned in the methodology section. Three hundred and sixty-three participants filled the online questionnaires presented. The initial population size was supposed to be equal between the two genders, but more male participants than the females filled the online questionnaires. The total N was three hundred and sixty-two participants.

The questions used in the questionnaires were converted into the variables needed to test the hypothesis. For instance the question "if you see a new game being broadcast by a large number of streamers you will?" the answers that followed included "more interested and by it," "decide whether to it," "will not buy it" and finally "repel the game." Each of the answers was given a numerical value from one to four; the first answer which was buying it was given the highest value among the others, which was four. The other questions were also used as variables, and the answers awarded numerical values to conduct quantitative research.

The reliability of the data had to be tested before analyzing it to present the results that prove the hypothesis is true or false. To do this, a test was done using Cronbach's alpha, and the results can be seen in the figure below.

Table 1cronbachs alpha

Cronbach's Alpha

Cronbach's Alpha Based on Standardized Items

N of Items

.902

.908

16

All the sixteen variables and their data set were tested for reliability using Cronbach's alpha and returned a value of .908 and .902; this means that the data is consistent and reliable. Because the rules of the Cronbach's alpha state that a Cronbach's alpha of above.70 means that the data is reliable and consistent. Meaning the results produced will also be accurate and can be trusted.

4.1 Descriptive statistics

A detailed descriptive statistics was done for each of the variables used. The mean, mode variance, and other statistical information were calculated for all the three hundred and sixty-two samples. Each of the variables used showed a 95% confidence level in layman terms; this means that ninety-five out of a hundred samples will contain the true mean; this also means that the data is reliable; hence the results produced can be trusted.

In terms of passing out the questionnaires, three main things were taken into consideration the age, gender, and education level of the participants. Out of the N-363, two hundred and twenty-six participants made, and the rest were females. The research's original intention was to have an equal number of males and females taking part in the study. But more males agreed to take part in the study as compared to the females. The data's central tendency was 1.37, meaning more males were closer to the mean than the females.

Many young people tend to use video games more than the other age groups, and the questionnaires required the participants to input their age details to confirm if young people play a lot of video games. For consent purposes, the youngest persons allowed to participate in this research were eighteen-year-olds, and the eldest was around fifty-one years old. The participants aged between eighteen and twenty-five were more than the others; to be specific, around six, over ten of the participants were between eighteen and twenty-five. The mean age was around 1.834, with a standard error of 0.0624.

The study also considered the participant's highest level of education to see whether there is any relationship between the two. Highly educated individuals will usually tend to perform a cost-benefit analysis before purchasing any product. At the same time, the individuals who haven't gone through a higher education system will purchase a product because of other things like their liking for the product. Five out of ten individuals who took part in this study have a degree or higher, meaning that five over ten of the individuals haven't gone through higher education systems. This means that the sample population is evenly distributed and is a good representation of the region under study.

4.2 Statistical Analysis

To determine if the hypothesis mentioned proved to be true or null, a correlation and regression analysis were done using SPSS.

The first hypothesis test was to determine if the live broadcasting of a game will affect the sales, and if the number of channels the viewer sees a certain game affects their intention of purchasing that particular game.

Table 2 correlation for hypothesis 1

Game sales

If you see a new game being broadcast live by a large number of streamers, you will?

Game sales

Pearson Correlation

Sig. (1-tailed)

N

1

363

.818**

.000

363

If you see a new game being broadcast live by a large number of streamers, you will?

Pearson Correlation

Sig. (1-tailed)

.818**

.000

1

N

363

363

The figure above presents the correlation results when the variable "if you see a new game being broadcasted by a large number of streamers you will? This variable was the independent variable, and the game sales dependent on whether the viewer’s bought a game based on the number of times the game was broadcasted.

The correlation coefficient of the two variables was between 0.818, and 1.This results show that the two variables have a positive correlation. This means whenever there is an increase in the broadcast of a game, there is also an increase in in-game sales.

To further prove that the two variables have a positive correlation, a regression analysis must be done considering the same variables mentioned above.

Table 3regression best line of fit for hypothesis 1

The regression analysis also shows that whenever there is an increase in the values of the independent variable, as seen in the y-axis of the figure above, there is an equal increase in in-game sales. The line of best fit is an upward sloping line, as seen in the figure above, which also proves that the increase in in-game sales depends on the number of times a game is being broadcasted.

The second hypothesis tests whether the number of viewers and real-time comments of a game will have positive effects on game sales.

To find out the correlation between the two said variables stat had to be selected from the variable that shows “How often do you watch games every week?” this variable shows how many times in a week a viewer watches a game hence affects the overall number of viewers watching a game.

Table 4correlation for hypothesis 2

Game sales

How often do you watch live games every week?

Game sales

1

How often do you watch live games every week?

0.737720414

1

The correlation analysis presented correlation coefficients of between .7 and one; this means that the two variables are positively correlated. An increase in the number of people watching a game will lead to an increase in the sales of a particular game.

A regression analysis was also conducted to try and test if the hypothesis is true.

Table 5regression for hypothesis 2

The regression analysis conducted also proves that the more people watch a live game in a week, the more the game sales increase.

Hypothesis three tests, whether live streaming multiplayer online games would affect game sales compared to non-multiplayer online games. The first step would be to find the correlation and regression of non-multiplayer online games against game sales and compare it to the correlation and regression of multiplayer online games.

The first step was to find the correlation between multiplayer online games and the game's sales produced the following result.

Table 6correlation for hypothesis 3

Multiplayer online games

Game sales

Multiplayer online games

Pearson Correlation

Sig. (1-tailed)

N

1

363

.769

.000

362

Game sales

Pearson Correlation

Sig. (1-tailed)

.769

.000

1

N

363

363

The correlation coefficient is between .769 and one, meaning that there is a positive correlation between multiplayer online games and game sales. Meaning a live stream of a multiplayer online game anytime is broadcasted; the viewers have an urge to purchase the game leading to an increase in the game sales.

To prove the correlation results, a regression analysis was also done and produced the following results.

Table 7regression for hypothesis 3

Model

Unstandardized Coefficients

B

Standardized Coefficients

Std. Error

t

Beta

Sig.

1

(Constant)

Multiplayer online games

19.306

.951

.486

.126

.769

39.703

7.528

.000

.000

Model

Unstandardized Coefficients

B

Standardized Coefficients

Std. Error

t

Beta

Sig.

1

(Constant)

19.306

.486

39.703

.000

Multiplayer online games

.951

.126

.769

7.528

.000

The regression analysis produced a regression coefficient of .769; this means that graphically, the best line of fit will produce an upward-moving slope line. This means whenever a multiplayer online game is broadcasted, and the viewer will purchase the game leading to a significant increase in in-game sales.

The next step is to find out the correlation between all the other games and game sales. The results of the correlation are as seen in the figure below. Each game had its correlation against the game sales.

The average of all the coefficients was arithmetically calculated, and the average correlation coefficient of all the games against the game sales is .337. This means that there is a weak correlation between the broadcasting of other non-multiplayer online games and game sales.

Game sales

Role-playing games

Leisure games

Indie games

Action games

Strategy games

Sports racing games

Game sales

Pearson Correlation

Sig. (1-tailed)

N

1

362

.409

.000

362

.336

.000

362

.478

.000

362

.412

.000

362

.284

.000

362

.443

.000

362

Role-playing games

Pearson Correlation

Sig. (1-tailed)

N

.409

.000

362

1

362

-.042

.213

362

.020

.350

362

.004

.471

362

.173

.000

362

.110

.018

362

Leisure games

Pearson Correlation

Sig. (1-tailed)

N

.336

.000

362

-.042

.213

362

1

362

-.025

.318

362

.097

.033

362

-.069

.096

362

-.041

.220

362

Indie games

Pearson Correlation

Sig. (1-tailed)

N

.478

.000

362

.020

.350

362

-.025

.318

362

1

362

.018

.364

362

-.008

.442

362

.123

.010

362

Action games

Pearson Correlation

Sig. (1-tailed)

N

.412

.000

362

.004

.471

362

.097

.033

362

.018

.364

362

1

362

-.061

.123

362

.027

.307

362

Strategy games

Pearson Correlation

Sig. (1-tailed)

N

.284

.000

362

.173

.000

362

-.069

.096

362

-.008

.442

362

-.061

.123

362

1

362

-.007

.444

362

Sports racing games

Pearson Correlation

Sig. (1-tailed)

.443

.000

.110

.018

-.041

.220

.123

.010

.027

.307

-.007

.444

1

N

362

362

362

362

362

362

362

A regression analysis was also conducted, and the results produced a regression coefficient of .337. The low regression coefficient means that the non-multiplayer online games will have an insignificant effect on game sales.

This means that multiplayer online games have a more positive effect on game sales than non-multiplayer online games.

Hypothesis four tests whether the audience liking to a particular streamer will affect the willingness of the audience to purchase a game and return affect the game sales positively. The correlation analysis was conducted using the variables “if your favorite streamer is broadcasting a new game, what will you do?” and its correlation to the game sales.

Game sales

If your favorite streamer is broadcasting a new game, what would you do?

Game sales

Pearson Correlation

Sig. (2-tailed)

N

1

363

.871**

.000

363

If your favorite streamer is broadcasting a new game, what would you do?

Pearson Correlation

Sig. (2-tailed)

.871**

.000

1

N

363

363

The correlation coefficient produced was between .871 and one; this means that the two variables show a positive correlation. So when an audience sees their favorite streamer broadcasting a new game, they will have a higher likelihood of purchasing that particular game.

The regression analysis was also conducted to predict if the increase in in-game sales depends on the broadcasting of a game by a viewer's favorite streamer. The analysis shows an upward sloping line of best fit, as seen in the figure above. Meaning whenever a viewer sees a game by broadcast by a favorite streamer, they have the urge and willingness to buy that particular game; this, in turn, increases the overall game sales.

4.3 Discussion