October 15, 2023
In this analysis, we are going to be looking at various elements that effect the top downloaded applications in the IOS Apple Store. From this dataset I obtained it from kaggle.com:
https://www.kaggle.com/datasets/ramamet4/app-store-apple-data-set-10k-apps
According to the post: “This data set contains more than 7000 Apple iOS mobile application details. The data was extracted from the iTunes Search API at the Apple Inc website.”
## X id track_name size_bytes ## Min. : 1 Min. :2.817e+08 Length:7197 Min. :5.898e+05 ## 1st Qu.: 2090 1st Qu.:6.001e+08 Class :character 1st Qu.:4.692e+07 ## Median : 4380 Median :9.781e+08 Mode :character Median :9.715e+07 ## Mean : 4759 Mean :8.631e+08 Mean :1.991e+08 ## 3rd Qu.: 7223 3rd Qu.:1.082e+09 3rd Qu.:1.819e+08 ## Max. :11097 Max. :1.188e+09 Max. :4.026e+09 ## currency price rating_count_tot rating_count_ver ## Length:7197 Min. : 0.000 Min. : 0 Min. : 0.0 ## Class :character 1st Qu.: 0.000 1st Qu.: 28 1st Qu.: 1.0 ## Mode :character Median : 0.000 Median : 300 Median : 23.0 ## Mean : 1.726 Mean : 12893 Mean : 460.4 ## 3rd Qu.: 1.990 3rd Qu.: 2793 3rd Qu.: 140.0 ## Max. :299.990 Max. :2974676 Max. :177050.0 ## user_rating user_rating_ver ver cont_rating ## Min. :0.000 Min. :0.000 Length:7197 Length:7197 ## 1st Qu.:3.500 1st Qu.:2.500 Class :character Class :character ## Median :4.000 Median :4.000 Mode :character Mode :character ## Mean :3.527 Mean :3.254 ## 3rd Qu.:4.500 3rd Qu.:4.500 ## Max. :5.000 Max. :5.000 ## prime_genre sup_devices.num ipadSc_urls.num lang.num ## Length:7197 Min. : 9.00 Min. :0.000 Min. : 0.000 ## Class :character 1st Qu.:37.00 1st Qu.:3.000 1st Qu.: 1.000 ## Mode :character Median :37.00 Median :5.000 Median : 1.000 ## Mean :37.36 Mean :3.707 Mean : 5.435 ## 3rd Qu.:38.00 3rd Qu.:5.000 3rd Qu.: 8.000 ## Max. :47.00 Max. :5.000 Max. :75.000 ## vpp_lic ## Min. :0.0000 ## 1st Qu.:1.0000 ## Median :1.0000 ## Mean :0.9931 ## 3rd Qu.:1.0000 ## Max. :1.0000
Games is highest recorded application genre in the IOS Mobile Store.
With the Games genre clearly being the highest recorded genre in the IOS Mobile Store in 2017, there is a lot of different questions we can ask and see if there is a statisical pattern among the data set.
Let’s start by seeing how we can make this use RStudio to make this dataset:
Null Hypothesis (\(H_O\)): There is no difference between free and paid games \[ H_O: \mu_{\text{Free}} = \mu_{\text{Paid}} \]
Alternative Hypothesis (\(H_a\)): There is a difference between free and paid games. \[ H_a: \mu_{\text{Free}} \neq \mu_{\text{Paid}} \]
Where: \(\mu_{\text{Free}}\): user rating of free games \(\mu_{\text{Paid}}\): user rating of paid games
\[ t = \frac{\bar{X_1} - \bar{X_2}}{\sqrt{ \frac{s_1^2}{n_1} + \frac{s_2^2}{n_2} }} \]
Where:
\(n_1\) and \(n_2\) are the sample sizes of both average ratings of paid games and free games
\(\bar{X_1}\) and \(\bar{X_2}\) are the sample means of both average ratings of paid games and free games
\(s_1^2\) and \(s_2^2\) are the sample variances of both average rating of paid and free games
The t-statistc is used to determine to reject the null hypothesis or fail to reject the null hypothesis.
games <- data %>% filter(prime_genre == “Games”) %>% mutate(appType = ifelse(price == 0, “Free”, “Paid”))
user <- mean(games$user_rating)
ggplot(games, aes(x = appType, y = user, fill = appType)) + geom_boxplot() + labs(title = “User Rating of Free Games vs Paid Games”, x = “App Type”, y = “Average User Rating”) + theme_minimal() + scale_fill_manual(values = c(“Free” = “pink”, “Paid” = “green”))
t_test <- t.test(user_rating ~ appType, data = games)
t_test
With the given output of this boxplot: Paid apps have a higher median at 4.5 compared to free apps with a median of 4 with few outliers for both.
## ## Welch Two Sample t-test ## ## data: user_rating by appType ## t = -8.3759, df = 3859.8, p-value < 2.2e-16 ## alternative hypothesis: true difference in means between group Free and group Paid is not equal to 0 ## 95 percent confidence interval: ## -0.4645133 -0.2883000 ## sample estimates: ## mean in group Free mean in group Paid ## 3.528578 3.904984
t-Test Result: \(t = -8.3759, p < 0.0001\)
Since the t value is negative, this means that Free apps has a lower mean than Paid apps
Usually we compare the p value to 5%, but since the outputted p value is so small, there suggests strong evidence to reject the null hypothesis.
Conclusion: We reject the null hypothesis.
We are taking into account the price, user rating, and the rating count total to see if there is a correlation between the 3 elements:
## track_name price rating_count_tot user_rating ## 1 Fruit Ninja Classic 1.99 698516 4.5 ## 2 Clear Vision (17+) 0.99 541693 4.5 ## 3 Minecraft: Pocket Edition 6.99 522012 4.5 ## 4 Plants vs. Zombies 0.99 426463 5.0 ## 5 Doodle Jump 0.99 395261 4.5 ## 6 Draw Something 2.99 360974 4.5 ## 7 Infinity Blade 0.99 326482 5.0 ## 8 Geometry Dash 1.99 266440 5.0 ## 9 Tiny Wings 0.99 219418 4.5 ## 10 Traffic Rush 0.99 213092 3.5
A lot of these top high reviewed games have one thing in common:
Constant maintance checks for bugs
Consistent updates with new features
Thank you for listening!
Questions?