Chapter II- Literature Review
3.7 Reliability and Validity Analysis
3.7.2 Validity
Validity means the extent of correctness of the variables. It refers to the extent that the tests or other measurement instruments are capable of measuring characteristics or functions that the researchers want to measure. Achieving a high score on validity means the results of the tests show more real characteristics in the variables the examiner is trying to measure. Validity coefficients of intrinsic validity are just the square root of the reliability coefficient. The American Psychological Association (APA) published a book titled “The Standards for Educational and Psychological Testing”
in 1974 which generalized the testing of validity into the following three types:
1. Content Validity: Checking the operationalization against the relevant content domain for the construct. It can test the appropriateness of content validity by according to a set of processes of a measurement tool.
2. Criterion-Related Validity: Checking the performance of the operationalization against some criterion.
3. Construct Validity: Testing the level validity of some theoretical concept or trait, normally based on some concept as basis of building a constructed correlation.
The level of accuracy is then based on the accuracy of the theory itself.
The variables examined in the study are based mainly from scholars with tested theoretical concepts, and the construction and presentation of the questionnaire survey is assisted by advising professor, students, and various papers, theses, and journal articles on the subject matter. Thus, the goal is to enhance the level of validity of the study with the help of various literatures and sorting out opinions from specialized scholars, making this study appropriate in terms of level of validity.
Chapter IV Data Analysis
A total of 394 copies of valid samples are retrieved, and using SPSS 17.0 software, statistical analysis will be conducted according to previously stated goal of research and defined hypotheses.
4.1 Factor Analysis
4.1.1 Factor analysis on attractiveness of game design element 1st run
The factor analysis conducted on the research variables uses principal factor analysis (PFA) and varimax to perform orthogonal rotation. The rotated factor loading matrix is obtained when Eigenvalue is bigger than 1.
KMO and Bartlett’s test of sphericity is used when conducting the factor analysis. KMO, also known as Kaiser-Meyer-Olkin, is the test of the adequacy of the samples as the tools of measurement for a study. According to Kaiser, when KMO >0.9 (excellent), KMO >0.8 (great), KMO >0.7 (Good), KMO >0.6 (Average), KMO >0.5 (Poor), and KMO <0.5 (Reject).
From Barlett’s test of shpericity, we can determine whether the data fits as a multivariate normal distribution, and that it can also be used to test whether the correlation coefficient matrix is fitted in order to proceed to the factor analysis. This study’s approximate chi-square is 9069.427, with a degree of freedom of 300 when the significance level is set at α = 0.05. In other words, the test is significant and that the correlation coefficient matrix representing the parent group does have an existing common factor. Furthermore, the KMO value is 0.929, meaning the study’s attractiveness of game design element is a fitted research variable to run the factor analysis.
According to Wu & Lin (2001), when factor loading is bigger than 0.6, it means the factor is significant, and we should reject the factor when it is less than 0.6.
Research conducted by Zaltman & Burger (1975) pointed out that when cumulative explained variance is higher than 40%, the result is within a reasonable range. After conducting the factor analysis on the attractiveness of game design element , the cumulative explained variance is 76.052%, as shown in table 4-1.
Table 4-1 Attractiveness of Game Design Element Eigenvalue and Explained Variance (1st Run)
Factor Eigenvalue Explained Variance Cumulative Explained Variance
Factor 1 12.787 51.147% 51.147%
Factor 2 2.156 8.25% 59.772%
Factor 3 1.730 8.920% 66.692%
Factor 4 1.281 5.122% 71.815%
Factor 5 1.059 4.238% 76.052%
Here is the result for the first run on the factor analysis (table 4-2):
Table 4-2 Rotated Factor Loading for Attractiveness of Game Design Elements 1st Run
Rotated Component Matrixa
Component
1 2 3 4 5
Story #02 .859 .145 .141 .084 .194
Story #01 .846 .203 .029 .014 .111
Story #04 .737 .255 .193 .158 .115
Story #03 .728 .189 .194 .265 .187
Story #05 .641 .237 .292 .282 .153
Sound & Music #04 .566 -.023 .318 .376 .262
Sound & Music #03 .502 .019 .481 .279 .345
Visual Presentation #04 .214 .749 .297 .214 .292
Visual Presentation #03 .156 .738 .317 .334 .177
Visual Presentation #02 .125 .728 .208 .298 .221
Visual Presentation #05 .295 .722 .233 .130 .244
Visual Presentation #01 .245 .714 .259 .130 .335
Interaction #01 .288 .309 .766 .145 .090
Interaction #04 .228 .325 .758 .067 .163
Interaction #02 .186 .444 .748 .186 .167
Interaction #03 .081 .391 .652 .302 .208
Control #01 .204 .238 .096 .842 .065
Control #03 .113 .234 .179 .819 .139
Control #02 .166 .363 .104 .814 .111
Sound & Music #02 .372 .006 .437 .555 .346
Sound & Music #01 .414 .047 .432 .537 .380
Character Setting #03 .047 .290 .219 .258 .736
Character Setting #02 .287 .364 .247 .142 .735
Character Setting #04 .361 .227 .039 .028 .702
Character Setting #01 .258 .441 .202 .178 .682
Extraction Method: Principal Component Analysis.
Rotation Method: Varimax with Kaiser Normalization.
a. Rotation converged in 8 iterations.
As shown in table 4-2, the number of factors is reduced from a total of six to five.
The “sound & music” factor is rejected and ruled out since all of the dimensions within the factor received a factor loading of less than 0.6, making them not significant for the test.
4.1.2 Factor analysis on attractiveness of game design element final run
After taking out the factor “sound & music” which received a factor loading of less than 0.6 for all its dimensions, a second and final run of factor analysis is performed, and the adjusted result shows an increase in the cumulative explained variance from 76.052% to 79.474%, as shown in table 4-3:
Table 4-3 Attractiveness of Game Design Element Eigenvalue and Explained Variance (Final Run)
Factor Eigenvalue Explained Variance Cumulative Explained Variance
Factor 1 10.830 51.571% 51.571%
Factor 2 2.003 9.540% 61.111%
Factor 3 1.575 7.501% 68.611%
Factor 4 1.270 6.046% 74.657%
Factor 5 1.012 4.817% 79.474%
And here is the result for the final (adjusted) run of the factor analysis (table 4-4):
Table 4-4 Rotated Factor Loading for Attractiveness of Game Design Element Final Run
Rotated Component Matrixa
Component
1 2 3 4 5
Visual Presentation #02 .787 .108 .194 .276 .196
Visual Presentation #01 .780 .226 .237 .104 .307
Visual Presentation #03 .756 .149 .325 .324 .172
Visual Presentation #05 .744 .286 .240 .115 .236
Visual Presentation #04 .739 .208 .321 .213 .300
Story #02 .152 .856 .133 .075 .198
Story #01 .152 .853 .058 .031 .141
Story #03 .102 .749 .238 .287 .234
Story #04 .248 .739 .204 .148 .126
Story #05 .168 .665 .324 .298 .188
Interaction #04 .232 .245 .806 .079 .205
Interaction #01 .277 .293 .773 .147 .104
Interaction #02 .406 .192 .767 .184 .184
Interaction #03 .246 .100 .730 .346 .278
Control #01 .154 .217 .151 .867 .116
Control #03 .180 .127 .208 .836 .171
Control #02 .328 .171 .133 .826 .137
Character Setting #03 .234 .060 .250 .258 .761
Character Setting #02 .353 .288 .250 .132 .741
Character Setting #04 .166 .359 .068 .050 .737
Character Setting #01 .402 .259 .229 .178 .703
Extraction Method: Principal Component Analysis.
Rotation Method: Varimax with Kaiser Normalization.
a. Rotation converged in 6 iterations.
An organized table with the factors, associated dimensions with their respective question contents and factor loadings is provided below (see table 4-5):
Table 4-5 Organized Table for Attractiveness of Game Design Element Constructs & Rotated Factor Loading
Factor Dimension Question Content Factor
Loading The character module design in the game is unique
and consistent. .787
The game delivers amazing style of visual arts. .780 The background design in the game is consistent. .756 The actions and expressions of the characters are
designed with great detail. .744
Factor 1
Visual Presentation
The equipments (armors & weapons) in the game have unique artistic designs that are visually
appealing. .739
The game offers a rich and intriguing story content. .856 The game delivers a storyline that clearly explains
the plot of the game. .853
The events in the game are consistent with one
another. .749
The game is expandable with lots of side stories &
quests to accomplish. .739
Factor 2 Story
The characters show development throughout the
progress of the game. .665
I am meet new friends easily in the game. .806 The players can conveniently team up with other
players for questing in the game. .773 The players can conveniently communicate with
other players in the game. .767
Factor 3 Interaction
The players can conveniently make in-game
transactions with other players in the game. .730 The control of the game is easy to handle. .867 The control of the game is easy to memorize. .836 Factor 4 Control
The control of the game is easy to learn. .826 The sets of equipments (weapons and armors) each
job class/race can wear are clearly different from
one another. .761
There are many techniques/ maigcs available to use
in the game. .741
The job classes/races are balanced in a way that no
one class/race is superior to another. .737 Factor 5 Character
Setting
There are many job classes/races to choose from in
the game. .703
4.2 Reliability Analysis on “Attractiveness of Game Design Element”, “Gamer Satisfaction” and “Gamer Loyalty”
According to Roberts and Wortzel, the alpha coefficient between 0.7 and 0.98 reflects high reliability. And if it is less than 0.35, then the researcher should give up on the variable(s) and look for other variable(s) to use.
Table 4-6: Reliability Score on Each of the Factor Constructs of the Attractiveness of Game Design Element
Attractiveness of Game Design Element Constructs
Cronbach’s α
Story (Q1,2,3,4,5) 0.905
Visual Presentation (Q6,7,8,9,10) 0.934 Character Setting (Q15,16,17,18) 0.888
Control (Q19,20,21) 0.910
Interaction (Q22,23,24,25) 0.915
Overall Construct 0.910
The coefficients are all within the 0.7 ~ 0.98 range, thus the chosen attractiveness of game design element items are highly reliable.
Table 4-7: Reliability Score on Gamer Satisfaction and Loyalty Gamer Satisfaction & Loyalty Cronbach’s α Gamer Satisfaction (Q26,27,28) 0.929 Gamer Loyalty (Q29,30,31,32) 0.931
The coefficients are all within the 0.7 ~ 0.98 range, thus the chosen satisfaction and loyalty items are highly reliable.
4.3 Analysis on Effects of “Demographic Variables” on “Attractiveness of Game Design Elements”
This section examines whether gamers with different “demographic variables” will have some influence towards their “attractiveness of game design element”. Therefore,
“demographic variables” including gender, age, and income will be set as independent
variables; “attractiveness of game design element” will be set as dependent variables.
Independent samples t-test analysis and one way ANOVA test will be used to examine these effects.
4.3.1 “Gender” and “attractiveness of game design element”
Testing hypothesis 1-1 H0: Gamers with different “gender” and attractiveness of game design element have no significant relationship.
Table 4-8 Independent Samples t-test for Gender on Attractiveness of Game Design Elements
Factor Dimension Gender Mean F value P value t-test for equality of means
Male 2.76 t p
Story Design
Quality Female 2.88
0.203 0.652
-0.921 0.358
Male 3.27 t p
Visual Presentation
Design element Female 3.46
0.858 0.355
At significance level α = 0.05, *when p < 0.05, significance is achieved
Hypothesis 1-1-1 H0: Gamers with different “gender” and attractiveness of game design element -“story” have no significant relationship.
Hypothesis 1-1-2 H0: Gamers with different “gender” and attractiveness of game design element -“visual presentation” have no significant relationship.
Hypothesis 1-1-4 H0: Gamers with different “gender” and attractiveness of game design element -“character settings” have no significant relationship.
Hypothesis 1-1-5 H0: Gamers with different “gender” and attractiveness of game design element -“control” have no significant relationship.
Hypothesis 1-1-6 H0: Gamers with different “gender” and attractiveness of game design element -“interaction” have no significant relationship.
This study uses independent samples t-test to examine the hypothesis, and the results found that under level of significance of α = 0.05, there are no significant relationships between gender and perceptions of quality of all the aspects of the attractiveness of game design elements- “story”, “visual presentation”, “sound & music”,
“character setting”, “control”, and “interaction”- thus we accept the null hypotheses 1-1-1H0,1-1-2H0, 1-1-4H0, 1-1-5H0, 1-1-6H0.
4.3.2 “Age” and “attractiveness of game design element”
Testing hypothesis 1-2 H0: Gamers with different “age” and attractiveness of game design elements have no significant relationship.
Table 4-9 One way ANOVA for Age on Attractiveness of Game Design Elements Age
Story 2.7587 2.7659 3.1077 3.0800 2.5000 0.639 0.635 Visual
Presentation 3.2550 3.3467 3.5231 3.7600 3.0000 0.568 0.686 Character
Setting 3.2690 3.2383 3.3173 3.8500 3.6250 0.425 0.791 Control 3.6924 3.6133 3.4872 3.1333 2.1667 0.847 0.496 Interaction 3.5071 3.4983 3.2308 2.7000 2.6250 1.160 0.328
At significance level α = 0.05, *when p < 0.05, significance is achieved
Hypothesis 1-2-1 H0: Gamers with different “age” and attractiveness of game design element -“story” have no significant relationship.
Hypothesis 1-2-2 H0: Gamers with different “age” and attractiveness of game design element -“visual presentation” have no significant relationship.
Hypothesis 1-2-4 H0: Gamers with different “age” and attractiveness of game design element -“character settings” have no significant relationship.
Hypothesis 1-2-5 H0: Gamers with different “age” and attractiveness of game design element -“control” have no significant relationship.
Hypothesis 1-2-6 H0: Gamers with different “age” and attractiveness of game design element -“interaction” have no significant relationship.
This study uses one-way ANOVA test to examine the hypothesis, and the results found that under level of significance of α = 0.05, there are no significant relationships between age and perceptions of quality of all the aspects of the attractiveness of game design elements- “story”, “visual presentation”, “sound & music”, “character setting”,
“control”, and “interaction”- thus we accept the null hypotheses 1-2-1H0,1-2-2H0, 1-2-4H0, 1-2-5H0, 1-2-6H0. Tukey HSD test, a post-event analysis was not needed since the p-value for all the factors are higher than 0.05 under α = 0.05.
4.3.3 “Monthly income” and “attractiveness of game design element”
Testing hypothesis 1-3 H0: Gamers with different “monthly income” and attractiveness of game design element have no significant relationship.
Hypothesis 1-3-1 H0: Gamers with different “monthly income” and attractiveness of game design element -“story” have no significant relationship.
Hypothesis 1-3-2 H0: Gamers with different “monthly income” and attractiveness of game design element -“visual presentation” have no significant relationship.
Hypothesis 1-3-4 H0: Gamers with different “monthly income” and attractiveness of game design element -“character settings” have no significant relationship.
Hypothesis 1-3-5 H0: Gamers with different “monthly income” and attractiveness of game design element -“control” have no significant relationship.
Hypothesis 1-3-6 H0: Gamers with different “monthly income” and attractiveness of game design element -“interaction” have no significant relationship.
This study uses one-way ANOVA test to examine the hypothesis, and the results (Table 4-10) found that under level of significance of α = 0.05, there are no significant relationships between age and perceptions of quality of these aspects of the attractiveness of game design elements- “story”, “sound & music”, “character setting”,
“control”, and “interaction”- thus we accept the null hypotheses 1-3-1H0, 1-3-4H0, 1-3-5H0. However, the factor “visual presentation” and “interaction” is found to be significant with a p-value of 0.026 and 0.045, respectively, a deeper analysis follows:
Table 4-10 One way ANOVA for Monthly Income on Attractiveness of Game Design Elements
At significance level α = 0.05, *when p < 0.05, significance is achieved Monthly Income (NTD)
Story 2.7336 2.8917 3.0462 3.9000 1.282 0.280
Visual
Presentation 3.3590 3.0729 4.0308 3.6000 3.115 0.026* 3>2 Character
Setting 3.2606 3.2865 3.3846 3.0000 0.097 0.962 Control 3.5819 3.5278 3.8974 3.8333 0.353 0.787
Interaction 3.5186 3.2396 4.0769 3.8750 2.705 0.045* 3>2
As a post-event analysis, Tukey HSD test is used to further examine whether the significance is truly significant. We see that using the Tukey HSD test, “visual presentation” is indeed significant to demographic variable of income at p = 0.032 for income groups (2) and (3) with level of significance of α = 0.05 (table 4-11). The level of perception of the gamers falling in the income group $20,001~$30,000 is significantly higher than the gamers falling the income group $10,001~$20,000. Thus we reject the null hypothesis 1-3-2H0. We see that “visual presentation” have a higher influence on gamers in the income group $20,001~$30,000, then the group $30,001~$40,000 and those with less than $10,000.
Table 4-11 Tukey HSD Test for Visual Presentation and Monthly Income Tukey HSD Test
$10,001~$20,000 0.28609 0.13927 0.170
$20,001~$30,000 -0.67176 0.33446 0.187 Less than $10,000
$30,001~$40,000 -0.24099 0.83672 0.992 Less than $10,000 -0.28609 0.13927 0.170
$20,001~$30,000 -0.95785* 0.34847 0.032*
$10,001~$20,000
$30,001~$40,000 -0.52708 0.84242 0.924 Less than $10,000 0.67176 0.33446 0.187
Visual Presentation
$20,001~$30,000 $10,001~$20,000 0.95785* 0.34847 0.032*
$30,001~$40,000 0.43077 0.89562 0.963 Less than $10,000 0.24099 0.83672 0.992
$10,001~$20,000 0.52708 0.84242 0.924
$30,001~$40,000
$20,001~$30,000 -0.43077 0.89562 0.963
When Tukey HSD test is performed on the “interaction” to see whether its 0.045 (very closer to the borderline of 0.05) significant is truly significant. This post-event analysis finds out that the groups $10,001~$20,000 and $20,001~$30,000 do have some sort of influence on one another due to the significance of 0.048, thus the Tukey HSD test suggests that these two income groups do have an significant relationship between one another (table 4-12). Thus, we reject hypothesis 1-3-6H0.
Table 4-12 Tukey HSD Test for Visual Presentation and Interaction Tukey HSD Test
$10,001~$20,000 0.27897 0.13676 0.175
$20,001~$30,000 -0.55837 0.32844 0.325 Less than $10,000
$30,001~$40,000 -0.35645 0.82167 0.973 Less than $10,000 -0.27897 0.13676 0.175
$20,001~$30,000 -0.83734 0.34221 0.048*
$10,001~$20,000
$30,001~$40,000 -0.63542 0.8726 0.869 Less than $10,000 0.55837 0.32844 0.325
$10,001~$20,000 0.83734 0.34221 0.048*
$20,001~$30,000
$30,001~$40,000 0.20192 0.87951 0.996 Less than $10,000 0.35645 0.82167 0.973
$10,001~$20,000 0.63542 0.82726 0.869 Interaction
$30,001~$40,000
$20,001~$30,000 -0.20192 0.87951 0.996
4.4 Analysis on Effects of “Demographic Variables” on “Gamer Satisfaction”
Testing hypothesis 2 H0: Gamers with different “demographic variables” and gamer satisfaction towards the game have no significant relationship.
Table 4-13 Independent Samples t-test for Gender on Gamer Satisfaction
At significance level α = 0.05, *when p < 0.05, significance is achieved
Table 4-14 One way ANOVA for Age on Gamer Satisfaction
At significance level α = 0.05, *when p < 0.05, significance is achieved.
Table 4-15 One way ANOVA for Monthly Income on Gamer Satisfaction
At significance level α = 0.05, *when p < 0.05, significance is achieved
Factor Dimension Gender Mean F value P value t-test for equality of means
Male 3.17 t P
Gamer Satisfaction Female 3.26 4.187 0.041 -0.585 0.559
Age
Hypothesis 2-1 H0: Gamers with different “gender” and the gamer satisfaction towards the game have no significant relationship.
Hypothesis 2-2 H0: Gamers with different “age” and the gamer satisfaction towards the game have no significant relationship.
Hypothesis 2-3 H0: Gamers with different “monthly income” and the gamer satisfaction towards the game have no significant relationship.
This study uses both independent samples t-test and one way ANOVA to examine the hypothesis, and the results found that under level of significance of α = 0.05, there are no significant relationships between gamer satisfaction and demographic variables “gender”, “age”, and “income”- thus we accept the null hypotheses 2-1H0, 2-2H0, 2-3H0.
4.5 Analysis on effects of “Demographic Variables” on “Gamer Loyalty”
Testing hypothesis 3 H0: Gamers with different “demographic variables” and gamer loyalty towards the game have no significant relationship.
Table 4-16 Independent samples t-test for Gender on Gamer Loyalty
At significance level α = 0.05, *when p < 0.05, significance is achieved
Table 4-17 One way ANOVA for Age on Gamer Satisfaction
At significance level α = 0.05, *when p < 0.05, significance is achieved.
Factor Dimension Gender Mean F value P value t-test for equality of means
Male 3.17 t P
Gamer Loyalty Female 3.16 1.223 0.269 0.060 0.953
Age (1)
17~23
(2) 24~30
(3) 31~37
(4) 38~44
(5)
>45
F Value p-value Turkey
Gamer
Loyalty 3.2133 3.0917 3.3365 3.0500 2.6250 0.401 0.808
Hypothesis 3-1 H0: Gamers with different “gender” and gamer loyalty towards the game have no significant relationship.
Hypothesis 3-2 H0: Gamers with different “age” and gamer loyalty towards the game have no significant relationship.
Hypothesis 3-3 H0: Gamers with different “monthly income” and gamer loyalty towards the game have no significant relationship.
Table 4-18 One way ANOVA for Monthly Income on Gamer Loyalty At significance level α = 0.05, *when p < 0.05, significance is achieved.
This study uses both independent samples t-test and one way ANOVA to examine the hypothesis, and the results found that under level of significance of α = 0.05, there are no significant relationships between gamer loyalty and demographic variables
“gender”, “age”, and “income”- thus we accept the null hypotheses 3-1H0,3-2H0, 3-3H0.
4.6 Analysis on Effects of “Attractiveness of Game Design Elements” on “Gamer Satisfaction”
Testing hypothesis 4 H0: Attractiveness of game design elements and gamer satisfaction towards the game have no significant relationship.
Hypothesis 4-6 H0: Attractiveness of game design element - “interaction” and gamer satisfaction towards the game have no significant relationship.
Hypothesis 4-1 H0: Attractiveness of game design element - “story” and gamer satisfaction towards the game have no significant relationship.
Hypothesis 4-2 H0: Attractiveness of game design element - “visual presentation” and gamer satisfaction towards the game have no significant relationship.
Hypothesis 4-4 H0: Attractiveness of game design element - “character settings” and gamer satisfaction towards the game have no significant relationship.
Hypothesis 4-5 H0: Attractiveness of game design element - “control” and gamer satisfaction towards the game have no significant relationship.
This study sets the 5 factors of attractiveness of game design element as independent variables, and gamer satisfaction as the dependent variable:
Y = a + bX1 + cX2 + dX3 + eX4 + fX5 + ε Y (dependent variable): gamer satisfaction
X1~X5 (independent variables): including the attractiveness of game design element such as story, visual presentation, character setting, control, and interaction.
a : intercept term
b~f : the regression coefficient for each of the independent variables.
ε : error .
Table 4-19 Regression Analysis (1)
ANOVAb
Model Sum of Squares df Mean Square F Sig.
Regression 426.128 5 85.226 203.137 .000a
Residual 162.784 388 .420
1
Total 588.913 393
a. Predictors: (Constant), Interaction Average, Control Average, Story Average, Character Setting Average, Visual Presentation Average
b. Dependent Variable: Satisfaction Average
Table 4-20 Regression Analysis for Attractiveness of Game Design Elements on Gamer Satisfaction
Coefficientsa
Unstandardized Coefficients
Standardized Coefficients
Model B Std. Error Beta t Sig.
(Constant) -.211 .120 -1.759 .079
Story Average .253 .037 .240 6.841 .000
Visual Presentation Average .212 .045 .206 4.743 .000
Character Setting Average .243 .043 .223 5.618 .000
Control Average -.041 .032 -.043 -1.279 .202
1
Interaction Average .389 .041 .370 9.385 .000
a. Dependent Variable: Satisfaction Average
* represents p < 0.05, + represents p < 0.1 R2 = 0.724, Adjusted R2 = 0.720
From table 4-19 above, we see that the predictive power of the perceived quality on attractiveness of game design elements on gamer satisfaction is R2 (adjusted) = 72.0%, making the appropriateness of the function very high, and that the regression effects is significant ( F(5,388) = 203.137, p = 0.00). From table 4-20, we can see that there are no significant effects from the perceived quality “control” on “gamer satisfaction” p = 0.202<0.05, thus we accept the null hypothesis 4-5H0. And that a F2P MMORPG’s gamer satisfaction is influenced by perceived quality on story, visual presentation, character setting, and interaction:
1. Story: test results show a t value of 6.841 and p-value is 0.000 < 0.05, thus we reject null hypothesis 4-1H0. This means F2P MMORPG gamers’ perceived quality on the game’s story design has a significant effect on gamer satisfaction.
The regression coefficient is positive 0.253, meaning the higher the gamer’s perceived quality on a game’s story design, the higher the gamer’s satisfaction towards the game, and vice versa.
2. Visual Presentation: test results show a t value of 4.743 and p-value is 0.000 <
0.05, thus we reject null hypothesis 4-2H0. This means F2P MMORPG gamers’
perceived quality on the game’s story design has a significant effect on gamer satisfaction. The regression coefficient is positive 0.212, meaning the higher the gamer’s perceived quality on a game’s story design, the higher the gamer’s satisfaction towards the game, and vice versa.
3. Character Setting: test results show a t value of 5.618 and p-value is 0.000 <
3. Character Setting: test results show a t value of 5.618 and p-value is 0.000 <