RPubs link information

• Rpubs link comes here: https://rpubs.com/Manali10/1046710

Introduction

• This dataset provides a comprehensive look at Airbnb prices in Amsterdam.
• The Airbnb platform has revolutionized the way people find accommodations while traveling, offering a wide range of options beyond traditional hotels.
• Because of its affordability, flexibility, and opportunity to see local neighbourhoods, Airbnb has grown in popularity, particularly in major cities.
• This dataset provides significant insights into the factors that influence Airbnb prices by taking into account a variety of factors such as room type, cleanliness and satisfaction ratings, number of bedrooms, distance from the city centre, and more.

Problem Statement

• The aim is to determine if there is a significant difference in guest satisfaction ratings between different room types.
• Outliers were identified and handled.
• descriptive statistics and hypothesis testing were used to summarise and investigate the distribution of guest satisfaction ratings for various room types.
• A QQ plot was used to compare the observed data to the expected quantiles of a normal distribution.

Data

• The dataset was obtained from the Kaggle website and the link is provided below. https://www.kaggle.com/datasets/thedevastator/airbnb-prices-in-european-cities?select=amsterdam_weekdays.csv
• Vriables:

1. realSum (The overall cost of the listing on Airbnb.)
2. room_type (the specific type of accommodation provided.)
3. room_shared (whether or not the room is shared.)
4. room_private (Whether or not the room is private)
5. person_capacity (the number of guests who can stay in the room together.)
6. host_is_superhost (Whether the host is a superhost or not)
7. multi (Whether there are multiple rooms listed or not.)
8. biz (Whether or not the listing is for commercial purposes.)
9. cleanliness_rating (The listing’s cleanliness score.)
10. guest_satisfaction_overall(The listing’s total guest satisfaction score)

Data Cont.

11. bedrooms (The listing’s number of bedrooms.)
12. dist (how far it is from the city centre. )
13. metro_dist (The distance from the closest metro stop. )
14. attr_index
15. attr_index_norm
16. rest_index
    
17. rest_index_norm
    
18. lng (the listing’s longitude.)
    
19. lat (the listing’s latitude.)

Data Pre-processing

*The read.csv() method was used to import the dataset, and the head() function was used to display the first five rows.

amsterdam_weekdays <- read_csv("/Users/manalika/Desktop/Applied Analytics/amsterdam_weekdays.csv")
knitr::kable(head(amsterdam_weekdays,5))

Data Pre-processing Cont.

*str() function is used to check the data type of the variables.

str(amsterdam_weekdays)

## spc_tbl_ [1,103 × 20] (S3: spec_tbl_df/tbl_df/tbl/data.frame)
##  $ ...1                      : num [1:1103] 0 1 2 3 4 5 6 7 8 9 ...
##  $ realSum                   : num [1:1103] 194 344 264 434 486 ...
##  $ room_type                 : chr [1:1103] "Private room" "Private room" "Private room" "Private room" ...
##  $ room_shared               : logi [1:1103] FALSE FALSE FALSE FALSE FALSE FALSE ...
##  $ room_private              : logi [1:1103] TRUE TRUE TRUE TRUE TRUE TRUE ...
##  $ person_capacity           : num [1:1103] 2 4 2 4 2 3 2 4 4 2 ...
##  $ host_is_superhost         : logi [1:1103] FALSE FALSE FALSE FALSE TRUE FALSE ...
##  $ multi                     : num [1:1103] 1 0 0 0 0 0 0 0 0 1 ...
##  $ biz                       : num [1:1103] 0 0 1 1 0 0 0 0 0 0 ...
##  $ cleanliness_rating        : num [1:1103] 10 8 9 9 10 8 10 10 9 10 ...
##  $ guest_satisfaction_overall: num [1:1103] 93 85 87 90 98 100 94 100 96 88 ...
##  $ bedrooms                  : num [1:1103] 1 1 1 2 1 2 1 3 2 1 ...
##  $ dist                      : num [1:1103] 5.023 0.488 5.748 0.385 0.545 ...
##  $ metro_dist                : num [1:1103] 2.539 0.239 3.652 0.44 0.319 ...
##  $ attr_index                : num [1:1103] 78.7 631.2 75.3 493.3 552.8 ...
##  $ attr_index_norm           : num [1:1103] 4.17 33.42 3.99 26.12 29.27 ...
##  $ rest_index                : num [1:1103] 98.3 837.3 95.4 875 815.3 ...
##  $ rest_index_norm           : num [1:1103] 6.85 58.34 6.65 60.97 56.81 ...
##  $ lng                       : num [1:1103] 4.91 4.9 4.98 4.89 4.9 ...
##  $ lat                       : num [1:1103] 52.4 52.4 52.4 52.4 52.4 ...
##  - attr(*, "spec")=
##   .. cols(
##   ..   ...1 = col_double(),
##   ..   realSum = col_double(),
##   ..   room_type = col_character(),
##   ..   room_shared = col_logical(),
##   ..   room_private = col_logical(),
##   ..   person_capacity = col_double(),
##   ..   host_is_superhost = col_logical(),
##   ..   multi = col_double(),
##   ..   biz = col_double(),
##   ..   cleanliness_rating = col_double(),
##   ..   guest_satisfaction_overall = col_double(),
##   ..   bedrooms = col_double(),
##   ..   dist = col_double(),
##   ..   metro_dist = col_double(),
##   ..   attr_index = col_double(),
##   ..   attr_index_norm = col_double(),
##   ..   rest_index = col_double(),
##   ..   rest_index_norm = col_double(),
##   ..   lng = col_double(),
##   ..   lat = col_double()
##   .. )
##  - attr(*, "problems")=<externalptr>

Data Pre-processing Cont.

• Data Type Conversion Using the lapply function and the factor function, we converted the specified columns in the amsterdam Weekdays data frame into factors.

columns_to_convert<- c('room_shared' ,'room_private','host_is_superhost','room_type')
amsterdam_weekdays[,columns_to_convert] <- lapply(amsterdam_weekdays[,columns_to_convert] , factor)

Descriptive Statistics and Visualisation

• The summary() function is used to find the summary statistics of all the variables in the dataframe.

summary(amsterdam_weekdays)

##       ...1           realSum                 room_type   room_shared 
##  Min.   :   0.0   Min.   : 128.9   Entire home/apt:538   FALSE:1097  
##  1st Qu.: 275.5   1st Qu.: 309.8   Private room   :559   TRUE :   6  
##  Median : 551.0   Median : 430.2   Shared room    :  6               
##  Mean   : 551.0   Mean   : 545.0                                     
##  3rd Qu.: 826.5   3rd Qu.: 657.3                                     
##  Max.   :1102.0   Max.   :7782.9                                     
##  room_private person_capacity host_is_superhost     multi       
##  FALSE:544    Min.   :2.000   FALSE:780         Min.   :0.0000  
##  TRUE :559    1st Qu.:2.000   TRUE :323         1st Qu.:0.0000  
##               Median :2.000                     Median :0.0000  
##               Mean   :2.792                     Mean   :0.3083  
##               3rd Qu.:4.000                     3rd Qu.:1.0000  
##               Max.   :6.000                     Max.   :1.0000  
##       biz         cleanliness_rating guest_satisfaction_overall    bedrooms    
##  Min.   :0.0000   Min.   : 4.000     Min.   : 20.00             Min.   :0.000  
##  1st Qu.:0.0000   1st Qu.: 9.000     1st Qu.: 92.00             1st Qu.:1.000  
##  Median :0.0000   Median :10.000     Median : 96.00             Median :1.000  
##  Mean   :0.1151   Mean   : 9.461     Mean   : 94.36             Mean   :1.283  
##  3rd Qu.:0.0000   3rd Qu.:10.000     3rd Qu.: 98.00             3rd Qu.:2.000  
##  Max.   :1.0000   Max.   :10.000     Max.   :100.00             Max.   :5.000  
##       dist            metro_dist        attr_index      attr_index_norm  
##  Min.   : 0.01506   Min.   :0.03653   Min.   :  40.93   Min.   :  2.167  
##  1st Qu.: 1.30206   1st Qu.:0.46298   1st Qu.: 127.91   1st Qu.:  6.773  
##  Median : 2.34137   Median :0.85601   Median : 208.18   Median : 11.023  
##  Mean   : 2.84162   Mean   :1.08944   Mean   : 271.01   Mean   : 14.350  
##  3rd Qu.: 3.64814   3rd Qu.:1.51063   3rd Qu.: 386.44   3rd Qu.: 20.462  
##  Max.   :11.18710   Max.   :4.41191   Max.   :1888.55   Max.   :100.000  
##    rest_index      rest_index_norm        lng             lat       
##  Min.   :  50.88   Min.   :  3.545   Min.   :4.776   Min.   :52.29  
##  1st Qu.: 163.47   1st Qu.: 11.391   1st Qu.:4.871   1st Qu.:52.35  
##  Median : 260.26   Median : 18.135   Median :4.890   Median :52.37  
##  Mean   : 341.54   Mean   : 23.799   Mean   :4.891   Mean   :52.36  
##  3rd Qu.: 469.29   3rd Qu.: 32.701   3rd Qu.:4.907   3rd Qu.:52.38  
##  Max.   :1435.10   Max.   :100.000   Max.   :5.011   Max.   :52.42

Decsriptive Statistics Cont.

• calculated the total number of missing values present in the dataframe.
• The outliers are located using the is.outlier function, and they are subsequently eliminated from the dataframe using the filter function and the %>% pipe operator.

sum(is.na(amsterdam_weekdays))

## [1] 0

is.outlier <- function(x) {
  lower_fence <- summary(x)[2] - 1.5 * IQR(x)
  upper_fence <- summary(x)[5] + 1.5 * IQR(x)
  x < lower_fence | x > upper_fence
}
outliers <- is.outlier(amsterdam_weekdays$guest_satisfaction_overall)
sum(outliers)

## [1] 48

amsterdam_weekdays <- amsterdam_weekdays %>% filter(!outliers)

Decsriptive Statistics Cont.

• creating a box plot to visualize the distribution of guest ratings across different room types in the dataset.

ggplot(amsterdam_weekdays, aes(x = room_type, y = guest_satisfaction_overall)) +
  geom_boxplot(fill = "steelblue") +
  labs(x = "Room Type", y = "Guest Rating") +
  theme_minimal()

Decsriptive Statistics Cont.

• Used group_by() to gain insights into the distribution and average guest satisfaction ratings across different types of rooms in Amsterdam.

amsterdam_weekdays %>%
  group_by(room_type) %>%
  summarise(
    Min = min(guest_satisfaction_overall, na.rm = TRUE),
    Q1 = quantile(guest_satisfaction_overall, probs = 0.25, na.rm = TRUE),
    Median = median(guest_satisfaction_overall, na.rm = TRUE),
    Q3 = quantile(guest_satisfaction_overall, probs = 0.75, na.rm = TRUE),
    Max = max(guest_satisfaction_overall, na.rm = TRUE),
    Mean = mean(guest_satisfaction_overall, na.rm = TRUE),
    SD = sd(guest_satisfaction_overall, na.rm = TRUE),
    n = n(),
    missing = sum(is.na(guest_satisfaction_overall))
  ) -> table1

knitr::kable(table1)

Hypthesis Testing

• Null hypothesis (H0): There is no significant difference in guest satisfaction ratings between different room types.

\[H_0: \mu_1 = \mu_2 \] * Alternative hypothesis (HA): There is a significant difference in guest satisfaction ratings between different room types.

\[H_A: \mu_1 \ne \mu_2\]

Hypthesis Testing - QQ Plot.

• ploted QQ plot and it shows that the data is approximately normally distributed.

qqPlot(amsterdam_weekdays$guest_satisfaction_overall,dist="norm")

## [1] 389 661

Levene’s test

leveneTest(guest_satisfaction_overall ~ room_type, data = amsterdam_weekdays) %>% as.data.frame()

• the Levene test compares the “guest_satisfaction_overall” variable across different levels of the “room_type” variable. Since the p-value (0.123258) is greater than the typical significance level of 0.05, we do not have enough evidence to reject the null hypothesis of equal variances. This suggests that the variance of the “guest_satisfaction_overall” ratings is approximately equal across the different room types.

Two-sample t-tests

We are performing two-sample t-tests to compare the guest satisfaction scores between different room types. As we had more than two variables in a column, we divided the data into three groups based on the room types: “private room,” “entire home/apt,” and “shared room.” Then, we are conducting pairwise t-tests between these groups.

private_room <- amsterdam_weekdays$guest_satisfaction_overall[amsterdam_weekdays$room_type == "Private room"]
entire_home <- amsterdam_weekdays$guest_satisfaction_overall[amsterdam_weekdays$room_type == "Entire home/apt"]
shared_room <- amsterdam_weekdays$guest_satisfaction_overall[amsterdam_weekdays$room_type == "Shared room"]
t.test(private_room, entire_home,var.equal = FALSE, alternative = "two.sided")

## 
##  Welch Two Sample t-test
## 
## data:  private_room and entire_home
## t = -4.8324, df = 1041.8, p-value = 1.552e-06
## alternative hypothesis: true difference in means is not equal to 0
## 95 percent confidence interval:
##  -1.7130881 -0.7236285
## sample estimates:
## mean of x mean of y 
##  94.61626  95.83462

t.test(private_room, shared_room,var.equal = FALSE, alternative = "two.sided")

## 
##  Welch Two Sample t-test
## 
## data:  private_room and shared_room
## t = 0.85694, df = 5.0802, p-value = 0.43
## alternative hypothesis: true difference in means is not equal to 0
## 95 percent confidence interval:
##  -3.540044  7.105891
## sample estimates:
## mean of x mean of y 
##  94.61626  92.83333

t.test(entire_home, shared_room,var.equal = FALSE, alternative = "two.sided")

## 
##  Welch Two Sample t-test
## 
## data:  entire_home and shared_room
## t = 1.4434, df = 5.0684, p-value = 0.2077
## alternative hypothesis: true difference in means is not equal to 0
## 95 percent confidence interval:
##  -2.322248  8.324812
## sample estimates:
## mean of x mean of y 
##  95.83462  92.83333

Discussion

• First, the dataset was checked for any missing values, and it was discovered that none exist for the relevant variables. The ‘guest_satisfaction_overall’ variable’s outliers were identified and eliminated from the dataset.
• Descriptive statistics and visualisations were created to better understand the connection between guest satisfaction and room type. To display the distribution of guest ratings among various room categories, a box plot was made. It is clear that the ‘Entire home/apt’ and ‘Private room’ categories receive the majority of visitor reviews.
• The Levene’s test was used to see whether the variances between the groups were equal. Due to the Levene’s test result’s p-value (0.123258) exceeding the significance level of 0.05, it was clear that the assumption of equal variances was accurate.
• We performed two-sample t-tests to compare the guest satisfaction scores between different room types. We split the data into three groups depending on the types of rooms because we had more than two variables in each column: “private room,” “entire home/apt,” and “shared room.” Performed pairwise t-tests between these groups after that.

Discussion cont.

• limitations:
• The analysis that is offered in this study has certain limitations. First of all, it ignores other potential elements that can affect customer satisfaction in favour of concentrating only on a single dataset. Second, the dataset might not provide complete information since it might not have included every relevant factor.
• future investigations:
• We can Consider including additional elements into the analysis, such as price, amenities, and location, to investigate their impact on guest satisfaction.
• We can Conduct a longitudinal study to look at variations in visitor satisfaction over time and see if there are any seasonal patterns.
• Investigate the impact of guest reviews and ratings on guest satisfaction, as they can provide us a useful information about the overall guest experience.

References

• www.kaggle.com. (n.d.). Airbnb prices in Amsterdam. [online] Available at: https://www.kaggle.com/datasets/thedevastator/airbnb-prices-in-european-cities