Who Plays Video Games?

Saumitra Sahi

Introduction

• In U.C. Berkeley, some kinds in the statistics class uses an alternative method to learning statistics and probability. They use computer labs, which for some students feel like an educational video game.

• The committee that designed the lab created a survey to determine the extent to which the students play video games and which aspects of video games they find most and least fun.

Questionnaire• How much time did you spend last week playing video and/or computer games?• Do you like to play video and/or computer games?• What types of games do you play?• Why do you play the games you checked above?• Where do you usually play video/computer games?• How often do you play?• Do you still find time to play when you’re busy?• Do you think video games are educational?• What don’t you like about video game playing?• Sex?• Age?• When you were in high school was there a computer in your home?• What do you think of math?• How many hours a week do you work for pay?• Do you own a PC? Does it have a CD-Rom?• Do you have an e-mail account?• What grade do you expect in this class?

The Data

• A simple random sample of 95 from the Statistics 2, Section 1 during Fall 1994 class of 314 students was taken. 91 of the 95 responded.

• The purpose of the questionnaire was to find out– The extent to which the students play video games– Which aspects the students find most fun– Which aspects the students find least fun

Frequency of Play

Type Percent

Action 50

Adventure 28

Simulation 17

Sports 39

Strategy 63

• According to the above table, at least 63% of the students play games

What types of games do you play?

Frequency of Play

• The table shows the amount of time put into playing video games a week before the questionnaire.

• Most people do not play video games in the course.

Time Count Bootstrap

0 57 197

.1 1 3

.5 5 17

1 5 17

1.5 1 4

2 14 48

3 3 11

4 1 3

5 1 4

14 2 7

30 1 3

Total 91 314

Frequency of Play

• Though the majority of students do not play games, there is a rough normal distribution among those that actually played video games, with a mean at about 2 hours (denoted by 6 on the histogram)

Distribution of time

0

10

20

30

40

50

60

1 2 3 4 5 6 7 8 9 10 11

Time

Co

un

t

Series1

Why do you play the games you checked?

Aspects they like

Why? Percent

Graphics/Realism 26

Relaxation 66

Eye/Hand coordination

5

Mental Challenge 24

Feeling of Mastery 28

Bored 27

• Most people play video games for relaxation

Aspects they dislike

• Many people feel video games take too much time, so they do not like to play it.

Dislikes Percent

Too much time 48

Frustrating 26

Lonely 6

Too many rules 19

Costs too much 40

Boring 17

Friends don’t play 2

It’s pointless 33

What don’t you like about video games?

Conclusions

• The majority of students in the Statistics 2, Section 1 course in Fall ‘94 did not play video games.

• The majority of the students who did not like video games felt it was a waste of time.

• Those who did play video games averaged around two hours per week.

• The majority of those who like playing video games play for relaxation.

A Twist to the Study

• After seeing these results from the Statistics Course at UC Berkeley, I decided to do the same survey for the 4th Period Statistics Class and do a statistical inference (chi-square test) about the goodness of fit for the results to see if the UC Berkeley results from 1994 hold with PHS results from 2007.

Aspects we like

• Similar to U.C. Berkley’s results, relaxation is still the number one aspect students like about video games.

Why do you play the games you checked?

Why? Percent

Graphics/Realism 31

Relaxation 80

Eye/Hand coordination

15

Mental Challenge 38

Feeling of Mastery 31

Bored 54

What we are expected to like• According to what the UC

Berkeley students like about video games, if our class has the same preferences, our percentages would be expected to be these.

• CAUTION: Since we are about to embark upon a goodness of fit test, we need to be sure that all individual counts (percentages in this case) are at least 1 and no more than 20% of the expected counts are less than 5. – Since these conditions are

met, we can continue.

Why? Percent

Graphics/Realism 36

Relaxation 92

Eye/Hand coordination

7

Mental Challenge 33

Feeling of Mastery 39

Bored 37

Hypotheses

• When doing a goodness of fit chi-square test, we have two hypotheses– Null Hypothesis (H0) : The proportions are still

the same– Alternate Hypothesis (Ha): At least one

proportion is wrong.

Test Statistic

• In order to obtain a test statistic which should aid us in determining if the proportions still hold we need to plug our results into this formula, where Oi is the observed value, and Ei is the expected value.

• After this calculation, we get the chi-square value to be 21.61. Now we are ready to obtain the P-value

P-value

• The P-value needs to be less that an a-value. Let us have 95% confidence in this calculation, therefore we will have an a-value of 5%. Using the test statistic, we calculate a P-value.

• Only if the P-value < a-value can we reject H0 and accept Ha. Otherwise we fail to reject it.

• We can easily calculate the P-value in this case using a calculator. We use the x2cdf( function, with the desired syntax of (chi-square test statstic, infinity, df). The df is the degrees of freedom which is simply the number of categories minus one. In this case, df = 5.

• Our P-value turns out to be 0.0006204 which is less than our a-vaule of 0.05, rejecting H0 and accepting Ha.

Aspects we dislike

• Unlike U.C. Berkeley’s class, the major con to playing video games was that it was frustrating to play.

What don’t you like about video games?

Dislikes Percent

Too much time 31

Frustrating 54

Lonely 15

Too many rules 15

Costs too much 38

Boring 15

Friends don’t play 15

It’s pointless 23

Goodness of fit test

• We run through a similar test for the aspects we dislike. With all conditions met, we run through the test with 95% confidence, a chi-square value of 137.65, and df of 7.

• We obtain a P-value of 5.6 E-28, which is practically zero. This is less than the a-value of 0.05, rejecting H0 and accepting Ha.