Click the Original, Code and Reconstruction tabs to read about the issues and how they were fixed.

Original


Source: Tableau/Axios.


Objective

The main objective of this data visualization is to highlight the inefficiency of U.S President Donald trump.This data is mostly targeted at political parties who wants to highlight the problem in Trump leadership. It conveys a message that President Donald Trump does not take his responsibilities seriously and has a very casual approach towards his Presidency. This data is basically a leaked white house schedule from Nov 2018 to Feb 2019. In this leaked schedule we see that a there is a entry called “Executive time” in his routine. According to the data sources Trump largely spends this time in his residence playing golf, watching television, reading newspaper etc. This data was published on Tableau (https://public.tableau.com/profile/ryansoares#!/vizhome/TrumpsExecutiveTime_2/Dashboard2) originally which started an debate that the US president spends 60% of his time doing nothing productive. Trump government backed this by saying that executive time is the time president spends on important productive calls and meeting. This was countered by asking simple questions like why these meetings and calls are not listed or documented in any way. President’s former chief of staff John Kelly confirmed that 60% of the president schedule is listed as “executive time”. Further reading of this debates and controversies is available on various site available online.

The visualisation chosen had the following three main issues


Issue 1
Ethnicity:- As stated above this data is from a leaked source, this raises various question on the ethnicity of this data, this data was published online but the source of the data is still not very clear. There are no documents to back this data. We are not sure if this data is even true or not. Some may choose to believe this data and some may not as there are no ethnecity.
Issue 2
Data Integrity:-This data shows the daily routine of President Trump from Nov 2018 to Feb 2019 but there are many days which are not listed and there is no data available for those days. No event or schedule is listed for these days and we cannot assume holidays on these days as some of this missing periods are in continuation of almost 10days.
Issue 3
Deception:- This data is recorded over the time period of 8:00am to 5:00 pm, there are no record of data before 8:00 am and after 5:00pm. The work of a Country’s president is almost 24hrs and his schedule is designed right from the time he opens his eys to the time he closes it, one can almost say that this data was only focused between the time period where executive timings were listed in majority. However there are someday where data does not fit into this scaling period of 8:00 am to 5:00pm and all these days have executive timings listed on them. It is likely to say that this data shows extended routine of the president only for those days where executive timing are listed and not for all the days. This is a way of deceiving the target audience by showing them the data as per convenience and not the complete set.


Solutions
In my visualization I have grouped the total timings listed as “executive timing” in a weekly order and we see that week12 was the week where he listed “executive timings” the most on his schedule. On checking the original data we see that it was the dated from 21-01-2019(Mon) to 25-01-19(Sat). From this we can track the news and see the main events happening in America at that time if we find that there was nothing important going on, we can guess that executive timings are just a way of listing time which was wasted or are of no importance. If it turns out to be other way we can guess that these timings are of actual importance, in both ways these format provide a better way to look at our data which is focused on executive timings listed by trump administration.

Reference

Ryan sores.(2019).Trump’s Executive time * retrived on May 7,2019 from Tableau website: https://public.tableau.com/profile/ryansoares#!/vizhome/TrumpsExecutiveTime_2/Dashboard2

Code

The following code was used to fix the issues identified in the original.

library(ggplot2)
library(readr)
library(tidyr)
library(dplyr) 
library(aweek)

## Importing data
library(readr)
trump <- read_csv("Axios _ President Donald Trump Private Schedules, Nov. 7, 2018 to Feb. 2, 2019 - data.csv")
head(trump)
## # A tibble: 6 x 11
##    week date  time_start time_end duration listed_title top_category
##   <dbl> <chr> <time>     <time>      <dbl> <chr>        <chr>       
## 1     1 07-1~ 08:00      11:00         3   Executive t~ executive_t~
## 2     1 07-1~ 11:00      11:30         0.5 Meeting wit~ meeting     
## 3     1 07-1~ 11:30      12:30         1   Executive t~ executive_t~
## 4     1 07-1~ 12:30      13:30         1   Lunch        lunch       
## 5     1 07-1~ 13:30      17:00         3.5 Executive t~ executive_t~
## 6     1 08-1~ 08:00      09:30         1.5 Executive t~ executive_t~
## # ... with 4 more variables: listed_location <chr>,
## #   listed_project_officer <chr>, detail_category <chr>, notes <chr>
trump1<-trump[ ,c(1,5,7)]
head(trump1)
## # A tibble: 6 x 3
##    week duration top_category  
##   <dbl>    <dbl> <chr>         
## 1     1      3   executive_time
## 2     1      0.5 meeting       
## 3     1      1   executive_time
## 4     1      1   lunch         
## 5     1      3.5 executive_time
## 6     1      1.5 executive_time
t_exe<- trump1 %>% filter(top_category == "executive_time")
head(t_exe)
## # A tibble: 6 x 3
##    week duration top_category  
##   <dbl>    <dbl> <chr>         
## 1     1      3   executive_time
## 2     1      1   executive_time
## 3     1      3.5 executive_time
## 4     1      1.5 executive_time
## 5     1      1   executive_time
## 6     1      0.5 executive_time
##Data preprocessing

#week1
w1 <- t_exe %>% filter(week == 1)
ws1<-sum(w1$duration)
head(ws1)
## [1] 12
#week2
w2 <- t_exe %>% filter(week == 2)
ws2<-sum(w2$duration)
head(ws2)
## [1] 17.16667
#week3
w3 <- t_exe %>% filter(week == 3)
ws3<-sum(w3$duration)
head(ws3)
## [1] 14.25
#week4
w4 <- t_exe %>% filter(week == 4)
ws4<-sum(w4$duration)
head(ws4)
## [1] 24.91667
#week5
w5 <- t_exe %>% filter(week == 5)
ws5<-sum(w5$duration)
head(ws5)
## [1] 30.16667
#week6
w6 <- t_exe %>% filter(week == 6)
ws6<-sum(w6$duration)
head(ws6)
## [1] 33.5
#week7
w7 <- t_exe %>% filter(week == 7)
ws7<-sum(w7$duration)
head(ws7)
## [1] 31.25
#week8
w8 <- t_exe %>% filter(week == 8)
ws8<-sum(w8$duration)
head(ws8)
## [1] 0
#week9
w9 <- t_exe %>% filter(week == 9)
ws9<-sum(w9$duration)
head(ws9)
## [1] 18
#week10
w10 <- t_exe %>% filter(week == 10)
ws10<-sum(w10$duration)
head(ws10)
## [1] 29.08333
#week11
w11 <- t_exe %>% filter(week == 11)
ws11<-sum(w11$duration)
head(ws11)
## [1] 25.66667
#week12
w12 <- t_exe %>% filter(week == 12)
ws12<-sum(w12$duration)
head(ws12)
## [1] 33.75
#week13
w13 <- t_exe %>% filter(week == 13)
ws13<-sum(w13$duration)
head(ws13)
## [1] 27.5
##making matrice with new data
executive_timing<-c(12,17.16667,14.25,24.91667,30.16667,33.5,31.25,0,18,29.08333,25.66667,33.75,27.5)
executive_timing <- round(executive_timing,digits = 1 )
weeks<-c(1:13)
df_exe<-data.frame(weeks,executive_timing)
View(df_exe)

Data Reference

Ryan sores.(2019).Trump’s Executive time * retrived on May 7,2019 from Tableau website: https://public.tableau.com/profile/ryansoares#!/vizhome/TrumpsExecutiveTime_2/Dashboard2

Reconstruction

The following plot fixes the main issues in the original.

##Plot
p1 <- ggplot(data = df_exe, aes(x = reorder(weeks,executive_timing), y = executive_timing,fill=(executive_timing)))
p1 <- p1 + geom_bar(stat = "identity") + coord_flip() + geom_text(aes(label = paste(executive_timing, sep = ""))) 
p1

Data Reference

Ryan sores.(2019).Trump’s Executive time * retrived on May 7,2019 from Tableau website: https://public.tableau.com/profile/ryansoares#!/vizhome/TrumpsExecutiveTime_2/Dashboard2