Redesigning The Hindu Data Point Stories (2022)

[venkatrajam]

For this assignment, we'll use data stories from The Hindu Data Point.

Select a story that you like, study it carefully and redesign it. Specifically I want you to focus on understanding the data that powers the story, and how it is visually encoded to tell the intended story. Document your design process, capturing the following:

• What is the story the author is trying to tell?
• What the data he/she is using to tell the story? Describe its details -- type of data, extent of the data, dimensions of the data, gaps in the data, what data is essential and what is irrelevant.
• How is it encoded, problems with it and how you attempted to improve it.

You may choose to expand or curtail the scope of the data used in the story, or add an additional dataset to tell the story better. But do not deviate from the main intent of the original story. In other words, it is a redesign exercise, and hence I do not want you tell a different, unrelated story.

While you should provide a link to the original story, it might be useful to capture and display inline, appropriate parts of the original visualization, and your own design iterations to produce a coherent documentation.

For reference, take a look at what the previous batches (2019, 2020, 2021) did with this assignment.

[Abhinavbansal20]

Section 153A: Cases jump six-fold, only 1 in 5 convicted

Name: Abhinav Bansal 216330003, M.Des

Article Link: Section 153A: Cases jump six-fold, only 1 in 5 convicted

Story the Author is trying to tell:

The author wishes to bring to the attention of the readers the rise in crimes committed under Section 153A (promoting enmity on the ground of religion, race, and place of birth), which has risen 6 times from 2014 to 2020. Further, it also shows the negligence of the authorities towards such crimes. This is made evident by the data presented which shows that Section 153A has the lowest conviction rates (20.2%) and the second highest Police Pendency Rate (64.3%) amongst the different types of crimes mentioned.

The Authors provides the following visualizations to support their claim:

• Line Graph showing the rise in Crimes related to Section 153A from 2014 to 2020.
• Average rise in cases for every state from (2015-17) to (2018-20), a three-year period.
• Case Chargesheeting Rate of Various crimes in a Bar Chart
• Pendency rate of Different Crimes in a Bar Chart
• Conviction Rate of Different Crimes in a Bar Chart

It is clearly visible that the law enforcements are showing an apathetic attitude towards cases registered against Section 153A, since they have the lowest conviction rates, and the highest pendency rates.

Data Analysis: Graph 1

The data being used here is a Line Graph from 2014 to 2020, showing the number of cases registered against section 153A. However, this information is not made clear on the axis. On hovering, one gets redundant piece of information. (Incident Count, and Year), something which is visible on the graph itself. Although it gives a country-wise view, it is tough to identify the real factors behind this sudden change. A comparative state-wise breakdown would have given more depth to story to gain insights. Change in governance (of a particular state) and issues such as CAA (Citizenship Amendment Act) may have played a role.

Data Analysis: Graph 2

Here, the authors try to compare the increase in Incidents related to Section 153A within every individual state over two sets of 3-year each.

Organising Principle: The arrangement has been made in such a way that the state with the highest average comes out on the top. The amount of change has not been quantified adequately although the story mainly talks about this ‘Change’ in crime rate. This difference should have been made evident, which would have allowed us to identify some common patterns.

Further, the population of the state may also have an impact on this data, so having that proportion would have been beneficial. The ‘Plus’ symbol used for encoding should have been on the right-end, since semantically the increase is shown on the right side.

Data Analysis: Graphs 3,4,5

Here, the authors have presented 3 different bar charts which highlight interrelated problems namely- Case Charge Sheet Rate, Conviction Rate and Pendency Rate. A better approach might have been to encode these three pieces of information on a common graph, since that would have made the evidence stronger. Further, the Charge Sheet Rate information seems redundant, as it doesn't bring forward a strong point against Section 153A.

Gathering Data

To make the case stronger, I started by gathering the state-wise data over a period of 6 years (2014-2020) from NCRB reports. Further, the rate of change in crimes was also noted for every successive years. This gave an estimation of whether this increase is consistent throughout, or whether there is a dramatic increase at any interval. Following this, rankings based on Population, and this rate of increase were added to find a correlation between the two. The data was recorded on an excel sheet. (Decrease was marked in green; and the top 3 increases under every year was highlighted in Red)

Initial Observations:

• Presenting Data in %age change would be misleading in some cases, since an increase from 1 to 2 would amount to a 100% change.
• The number of cases is so small w.r.t the population of the states, that all of them lie in the bracket of (0.1 to 0.4), a very small interval to differentiate. Hence, absolute values would be considered.
• There seems to be a spike in the numbers of Assam in Delhi from 2019-2020.
• Despite being 7th and 12th in population ranking, Tamil Nadu and Telangana are 2nd and 3rd respectively in rate of Crimes Committed.
• Despite most of the country being in a lockdown, 2020 has seen a 80% rise in Section 153A crimes from 2019.

Redesigning

The story essentially comprises of 2 aspects:

A) Increase in Cases- shown by an overall line graph, and an aggregated increase of incidents per state.

Here, I feel the two graphs can be integrated to create one combined visualization. Some initial explorations were:

1. Individual Line plots of States against the Overall. However, too many lines make it difficult to read this graph in one go.
2. Grouped Bar Graphs: Again, too many elements and the main story is getting lost here.
3. States on X axis, Year on Y and plotting Tree Map based on only colour intensity. Line/ Dot plot can be used to plot overall data.
4. Further, ranked comparison of states can also be done. (histogram/ dot plots)

Explorations of Part A

B) Condition of Section 153A cases wrt other cases.

1. Charge Sheet Cases seems redundant
2. To make more impact on Section 153A, the two graphs can be combined. (Since 153A ranks lowest, and second lowest on them respectively)
3. Dot Plot used to mark the two. Since Pendency should be low and conviction high; more distance between the two plots would indicate ignorance by concerned authorities.

Explorations for Part B

Final Redesign- Iteration 1

[shweta-ratanpura]

While fake ₹500 notes double, conviction rates for counterfeiting cases remain around 30%

Shweta Ratanpura 216330016, M.Des

---

0.1 Number of counterfeit notes

Chart data converted to excel

Calculated Data

Conclusion

• Counterfeit 500 notes increase by 100% from '21 to '22
• Counterfeit 2000 notes increase by 50% from '21 to '22
• Drop in all counterfeit notes from '20 to '21 by 30% - likely due to pandemic

---

0.2 High Pendency

Chart data converted to excel

Calculated Data

Conclusion

• The article does not mention the drop in percent of investigations completed - There was a 26% drop from '19 to '20
• Data if for 2020 - not recent, while data for counterfeit notes is from 2022 - So how is The Hindu comparing the two? (See article heading)
• Pandemic hit in 2020 - which could be the reason for more pending cases

---

0.3 Poor Conviction

Chart data converted to excel

Calculated Data

Conclusion

• Rather than showing conviction percentage (30%), saying that 98% trials are pending in court as of 2020 is more convincing
• Although conviction percentage does not have a lot of change from past years, actual number of convicted cases has decreased significantly

  ---

1.1 Number of counterfeit notes - Explorations

Exploration 1

Data used

Data Viz created

View the interactive chart

Exploration 2

Data used

Data Viz created

View the interactive chart

Exploration 3

Data used

Data Viz created

View the interactive chart

Exploration 4

Data used

Data Viz created

View the interactive chart

Comparison

---

0.2 High Pendency Explorations

Exploration 1

Data used

Data Viz created

View the interactive chart

Exploration 2

Data used

Data Viz created

View the interactive chart

---

0.3 Poor Conviction Explorations

Exploration 1

Data used

Data Viz created

View the interactive chart

Exploration 2

Data used

Data Viz created

View the interactive chart

---

Final Data Viz Compilation

[ApoorvAnurag]

Apoorv Anurag 216330001 IxD M.des 2nd year

Article chosen

With poor pension rates and high health costs, is India ready for the next demographic phase?

Article published on 14th July 2022 By: Vignesh Radhakrishnan, Jasmin Nihalani

Content List:

1. Article's main theme
2. Data Viz in the article
3. Meaning of special terms used in the article
4. Things which can be improved in them
5. Basic data and thoughts which has to be visualized
6. Initial Ideations
7. Looking for data
8. Final Ideations
9. Rendering

The main theme of the article:

1. In the next 80 years share of the working population will decline and the share of dependent residents especially the elderly (above 65+ years of age) will increase.
2. In India, at the present time, a lot of the elderly don’t get pensions after retirement. If this trend continues they won’t be able to manage their expenses on their own and dependency on the younger generation will increase.
3. India has one of the highest shares of out-of-pocket expenditures on health. This is a twofold blow given the poor monetary support and sharp increases in health-related expenses post-retirement.

Data Viz provided in the article:

Meaning of special terms: To understand the story presented it was important for me to understand the demographic jargon used in the article and in the data visualizations. Some of them are explained below: Dependency ratio: The age dependency ratio is the ratio of dependents, i.e. people younger than 15 or older than 64, to the working-age population- those ages 15-64. Out-of-pocket expenditure ( as a % of current health expenditure): This indicator estimates how much are households in each country spending on health directly out of pocket. It estimates the share of out-of-pocket payments of total current health expenditures. Household out-of-pocket expenditure on health comprises cost-sharing, self-medication, and other expenditure paid directly by private households, irrespective of whether the contact with the health care system was established on referral or on the patient's own initiative.

Things which can be improved in these data viz:

Apart from these following things are also a pain point for the readers: Repeatedly using a scatterplot might confuse the reader between each compared data. Scatterplots use similar colors which makes them more confusing to look at. The data source (Ourworldindata, UN, International Labour Organization) provided at the bottom of the article did not contain all the data that were used to make the visualizations. Link: https://ourworldindata.org/

Initial Ideations

To overcome these unrequired pain points following ideations were made:

Looking for data

It was very difficult to gather some data such as percentage of elderly receiving pensions worldwide, etc. Some data was calculated from initially found data such as % the population share of the elderly in India. Some of the data that were compiled are as follows:

1. The population of India from 1950 to that expected in 2100. Source
2. Dependency Ratio of 65+ age group people in India. Source
3. Percentage share of population 20-64 age group and 65+ age group. Source
4. Median Age in India from 1950 to that expected in 2100. Source
5. Statewise 60+ percentage Distribution of Projected Population (2021-2036) by age-group in India. Source
6. Out-of-pocket as share of total expenditure (%)as % of current health expenditure in various countries, 2019. Source

Final Rendering

This plot shows % share of the population of the working class compared to the elderly population share. Only necessary trend lines are shown in this trendline plot, hence making it easy to comprehend.

Median age is the most used metric to show the approximate age of the overall population size. This trendline plot shows how the Indian citizen's median age is increasing over the years.

This parallel coordinate plot shows the statewide trend of elderly population share over the years 2021, 2031, and 2036. This gives a holistic picture of each state with the passage of time. Kerala and Tamil Nadu are supposed to be having largest share of elderly population in 2036. (Use the interactive prototype to see the value).

This bar chart shows Old-age pension beneficiaries as a proportion of the population above statutory pensionable age in 2012 among various major middle-income countries. India is shown with a pop effect in Red. The position of India is on the lower side of the graph.

Interactive Prototype Link

This choropleth map shows the elderly dependency ratio of all the countries in 2021 and compares it to that in 2100. We can easily see a darker color in 2100 among all the countries, showing an increase in the elderly dependency ratio over the years.

Link of the interactive choropleth

[aliviachaudhuri]

Tracing the link between dengue outbreak and monsoon season

Link to the Hindu article

Alivia Chaudhuri M.Des, 216330017

Authors intention:

From the news article, we can summarize that the author claims these main points:

1. There is a correlation between the pattern of dengue breakout along with the pattern of Indian monsoons across the states.
2. The burden is shifting every five years to the south.
3. In 2022, Andhra Pradesh, Karnataka and Tamil Nadu made up 60% of India’s cases.
4. Similar spikes were seen in 2012 and 2017.
5. Monthwise breakup reveals the dengue progression in the country mimics Indian monsoon - south-west and northeast monsoon winds.
6. There are some states like Maharashtra and Kerala whose number of deaths is particularly disproportionate to their share of cases since 2008.

> Critique of the current data viz:

• The first heatmap table looks incomplete. The pink base colour represents how much % cannot be accurately ascertained. The user has to make extra effort to go through other pink/red tiles and estimate that it may be below 9% (9% is one shade lighter than the lightest pink colour in the cells)
• Colour is not a good way to represent ratio data (percentage of dengue cases).
• The article claims there is a cycle every 5 years in the South, which is not highlighted in the table.
• “In 2022, Tamil Nadu, Karnataka and Andhra Pradesh formed 60% of India’s cases.” This claim is not supported by the data I found available on the website of[ National Centre for Vector Borne Disease Control. [DENGUE/DHF SITUATION IN INDIA])]((https://nvbdcp.gov.in/index4.php?lang=1&level=0&linkid=431&lid=3715). The number of total cases coming from Tamil Nadu, Karnataka and Andhra Pradesh makes up 49% of India’s cases.
• There is no way to verify the month-wise monsoon map. The size of the dots represents several cases so that we can compare them at a glance. However, since there is no legend attached to the maps, it is not possible to estimate the number of cases monthly.
• Dots in the dot map are opaque. Since they overlap, it is hard to estimate the number of cases and where precisely the place is located. The overlapping dots do not give any idea of the cases they obscured.
• It would be a good idea to map monsoon wind direction along with cases on India's map. The 2008-22 table can be shown better as we are comparing 2 quantities.

> Tabulating the data:

Number of dengue cases and deaths statewide from 2015-2022 (May 31, 2022)

Percentage of cases and deaths yearwise:

Average cases and deaths statewide with latitude and longitude:

Note: Since data for the month-wise dengue cases were not available, the part of the graphic containing the dot map in India has not been redesigned.

Attempt 1

Dot plot for the percentage of cases reported and percentage of deaths per state The chart looked complex and was not conveying the story to the user at a glance, though it represented all the attributes and items of the dataset.

Attempt 2

Radar chart showing cases for different states over the years. At first, I tried with the years as the circular axis, but the number of states and union territories was too large for plotting clearly. Exchanging the axes did not help either.

Attempt 3 Interactive map

Grouped column chart. The years are not visible clearly, and the grouping has been done statewise.

Grouped column chart by interchanging the axis.

I realised that the data for all the states over 8 years was becoming very complex to depict. So to get an overall sense of the data, I took the mean of total cases and deaths in India. Line chart

Another interactive line chart

Dot map

Final visualization It was unnecessary to show all 28 states and 8 UTs as not all the states got affected by dengue as severely as some. Hence I chose to show the 17 specific ones mentioned in the article. Interactive chart

Every 5 years, the higher rate of dengue virus shifts to the south of India. (as seen by the line graph of Tamil Nadu)

This trend is also seen in most of the Southern states

The next highest cases are from Punjab and West Bengal

To support the article's second point, I made a column chart to show the state's share of the national average of dengue cases and deaths. It shows the disproportionate share of Maharashtra's deaths, followed by Rajasthan and Kerala.

Interactive chart

[Tejaswinipundge]

Student- Tejaswini Pundge, 216330012 Ixd, IDC

Title- Quality of jobs will decline further as Agnipath scheme lacks social security benefits

About article

The article is about data collected by PLFS ( Periodic labor force survey) which is basically statistical surveys conducted in a number of countries designed to capture data about the labor market. The article aims to visualize the percentage of regular wage /salaried employees in India who lack social security benefits and lack pension or gratuity benefits.

Vulnerable job share

Table below shows the share of regular wage/ salaried employees who had no job contracts, were not eligible for paid leave and were not eligible for social security benefits in sectors other than agriculture. The table shows the percentage of rural, urban and total population for three different parameters in three different years.

Gender divide

Table below shows the percentage of male, female and total regular wage/salaried employees for no written contracts, No paid leaves and No social security benefits.

State-wise share

The graph below shows state-wise distribution of regular wage salaried employees with No job contracts, were not eligible for paid leaves, were not eligible for social security benefits and all of the above.

Problems :

Vulnerable job share and Gender divide chart- These are simple charts with a lot of numbers for different parameters and different years. Hence it is difficult to find information and compare two entities.

State-wise share graph - There is a lot of confusion about signs used for different parameters. Also, Signs are sufficient to differentiate between different parameters, Hence again using different colors for different parameters is not making sense. They are trying to keep colors consistent for each parameters, But difference in tints and adding one more parameter (i.e All of the above) confusing the user,

How my design is solving the problems

Converting the table data into graph For a data table of vulnerable job shares, a set bar graph is used to visualize the percentage of three categories i.e who don't have a written contract, don't have paid leaves, and don't have social security benefits at three different years i.e 17-18, 19-20, 20-21. Set bar graphs make it easy to read and compare.

For a data table of gender divide, a parallel coordinate graph is used in order to compare the data of three different parameters of three different categories. In both cases converting the data tables into graphs proves Weber’s law i.e We judge based on relative, Not absolute.

State wise share

[zuhaasif]

India reduced armed imports from Russia while china’s dependency increased

Article Link Name: Zuha Asif P M.Des, 216330014 In this article the author claims how India has reduced its dependency on Russia which was the major supplier until last 5 years. The data shows how India’s arms imports have significantly reduced in the last 5 years. Data from 1991 until 2017 shows a major dependency on Russia until the last drop from $15,356 million to $15,356 million TIV. (Trend Indicator Value) In 2021 France replaced Russia as India’s primary source, despite which Russia has fulfilled 46% of India’s defense needs in the last 5 years. The author has used 4 charts to support this article:

1. A Bar chart showing Total Imports by India over the years 1952 – 2021
2. A multiple line chart showing top 5 sources of India’s imports from 2000 – 2021
3. A scatter plot showing total arms imported from Russia between 2017 and 2021.
4. A scatter plot showing country’s dependency on Russia Graph 1 The chart shows India’s armed imports from Russia in 5-year intervals starting from 1952 to 2021. We can notice a significant increase from 1952 to 1991 with a huge drop in the year 1991. It dropped from $16,222 million to $5052 million, followed by a significant rise from the period 1992-2016. The graph ends with a drop from in the last 5 years, which the author is targeting. Since the period is 5 years interval, the data is a bit abstract to identify which year is the drop and the reason behind it. Graph 2 The second graph shows multiple line chart showing top 5 sources of India’s imports from 2000 – 2021. It clearly shows Russian imports were significant high when compared to other countries. The graph only gives a larger picture of Russia being a higher importer, but the graphs of United Kingdom, France, United States and Israel is of similar range and overlaps at the bottom making it unable to distinguish between the different graphs. The author also tries to depict that in 2021 France exceeds the Russian imports, which is not very evident in the bigger picture.

Graph 3 This is a scatter plot showing the dependencies of various countries on Russia in percentage on y axis. And the total arms imported from 2017-21. This is very difficult to find out which all countries are being shown and many of the plots tend to overlap each other. This graph only shows an overall distribution of dependencies on Russia. The dots which stand out are China which has dependence of 81.34 which has a total sum of $5337 million. Whereas India even though having imported 7068TIV has only 46% dependency on Russia.

Graph 4 Change in Dependency The chart shows change in a country’s dependency on Russia between 2012-16 and 2017-21 periods. The 2 different colours shows the positive and negative change. I Personally feel that scatter plot is not the best way to represent change. In this graph also India and china stands as the dependency on Russia decreased by 22% points between the two periods, whereas China’s dependency increased by 28% points.

Redesign

Explorations: As Initial line graph did not emphasize much on France as a major importer crossing Russia in the year 2021 an area chart better shows the transition. Still in the case the rest of the countries are not visible properly as there are many over laps in the same range. India's top 5 Sources Graph Since the data is continuous over a timeline, line chart would be the most apt. Since the visibility is an important factor Russia can be represented separately. The range of scale is also different in these cases. Since we are targeting on Russian imports the graphs can also be split into two, with Russia separate and the other major 4 exporters in a different graph. Here the scale on the y axis is different. Dependence Ratio Graph In the scatter plot diagram. As there were many overlaps it was very difficult to distinguish between different countries and what it was representing. In order to make this better, a Symbol map can be used wherein countries are plotted on a world map and the size indicates the dependence ratio and colour indicates the sum of Russian imports. From this map it is very evident that China has a stark contrast when it comes to dependence ratio.

Change in Dependency Graph

The change in dependency is shown in terms as a positive or negative change. When it is represented in scattered plot it is confusing. This can be instead represented as a bar charts, so that the positive and negative dependencies are on either sides and there is a clear demarcation. To show the sum of Russian imports I have incorporated line graph along with bar charts.

Redesigned Graphs

[ghost]

Mithun Murali M.Des, Roll 216330008

Enforced Vegetarianism in India

The article analyzed here is: How many Indians eat meat?

Overview

This article was published on the occasion of 2022 World Health Day, which falls on April 7th The source of this article is the data from the National Family Health Survey -5 (2019-2020). This article schematically proves that most of India's population consumes fish, chicken, and meat daily, weekly, or occasionally. The significance of this article is at the point when there are a lot of official and non-official drives and campaigns targeting vendors who sell meat that has been conducted in different parts of the country. The article shows that in over half of the States/ UTs analyzed, more than 90% of the population consumed fish or chicken, or meat daily or weekly, or occasionally. In 25 of them, the figure was more than 50%. In none of the States/UTs was the share less than 20%.

The stereotype of India as a vegetarian land is deep. However, National Family Health Survey -5 data released by the government has pricked the bubble of India as a vegetarian nation. As per this survey, only 20 % of Indians are vegetarian. Eighty percent are non-vegetarians: These include people who eat any combination of fish, meat, and eggs. The numbers have always shown India as a non-vegetarian nation – even as the stereotype persists.

Source data visualization

The data visualization done on the source article is a political map of India color coded between the Nonvegetarain majority states in saturated red, the vegetarian population majority states in saturated green, and the others with a darker shade of yellow and lighter shades of yellow accordingly. The visualization in the form of a map is helping in this context as the camouflaged purpose of this representation is to show the existing vote bank of right-winged political parties and also the states where they don't have much foothold. The meat ban or forced vegetarianism is intended to gradually convert the lost votes in the red-colored states over the cover of religion.

Extracting the data from the given visualization into tabular format:

The other side of this story

The ruling government has made food habits a part of its politics, with a slew of bans on the consumption of certain meat items in many parts of India.

India has five states that could be considered vegetarian (defined as having at least, half their population as vegetarian). These states are Rajasthan, Haryana, Gujarat, Madhya Pradesh, and Punjab. Remarkably, three out of these five have NDA chief ministers, and the other two recently lost the coalition.

The converse holds as well: India has sixteen states which are almost completely non-vegetarian, that is with less than 10% vegetarians.

In South India, the only state where the NDA has a significant presence is also, coincidentally, the only state where vegetarians have a considerable presence: Karnataka with 21% vegetarians.

Correlation and causation

Vegetarianism and the right-wing government have a relatively high correlation. Since 1996, the correlation coefficient of the proportion of vegetarians in a state and the number of NDA governments it has elected is relatively high.

Of course, correlation is not causation. Vegetarianism, in this case, is a mask for caste as well as geography. Unlike in the West, vegetarianism is rarely an individual choice in India, being dictated by identity. North and West India are far more inclined towards their veggies than the East or the South.

Thus, as it happens, vegetarian populations are the traditional vote banks of the NDA– which rather clearly explains the high correlation.

I tried to represent this data in different formats so that the same idea is conveyed.

Attempt 1: In this attempt, I used a Tornado Plot/diagram to represent this data. My focus was to show the comparison between each state's certain population and non-vegetarian population. Here the viewer can directly explore the stark contrast between these two target populations. Interactive Chart

Attempt 2: Here I used the staked column chart which has the same intention as attempt 1. Interactive Chart

Attempt 3: Pictorial human figure representation is used here for each state. . Interactive Chart

[rubayatahmed]

Name: Rubayat Ahmed 216330013, M.Des

Article : India’s press freedom ranking slips to 150, its lowest ever

Context : Reporters without Border is an international non-profit organization of journalists governed by principles of democratic governance, who do research every year and publish press freedom ranking of all the countries. This rankings reflects the state of journalism and right to information in a country, which is one of the main pillars of democracy. The ranking is based on a country’s performance in five broad categories political context, legal framework, economic context, sociocultural context and safety of journalists. India being the largest democracy in the world is expected to improve upon the press freedom and practice healthy journalism. But India India fell to the 150th position, its lowest ever as In the last edition, India was ranked eight positions higher, at 142. The author wants to highlight this degradation of rank in comparison to other major countries and also highlights the pattern of how in last few year the ranking has keep lowering.

Graphs : Author provides three visualization to support the article. These visualizations are based on data from Reporters without Border's official website.

1. Rankings of 180 countries in the 2022 Press Freedom Index
2. India’s rankings across various categories in 2022
3. India’s rank in the Press Freedom Index in last few years

Graph 1 : Rankings of 180 countries in the 2022 Press Freedom Index

• Not all the countries are shown
• Absence of top and bottom countries in each group makes less sense to form such group
• Because of the way each country represented as a line in each group, the weight of each group looks different though every group consist of 30 countries
• Color coding is not helping in any way, there is no meaning behind the colors. Rather it confuses viewer as if they have some distinct properties, but its not. It also overpowers other visual elements.

Graph 2 : India’s rankings across various categories in 2022

• To show just one single information the whole page has to be covered with redundant rankings(Y-axis)
• Comparisons could have been much more valuable

Graph 3 : India’s rank in the Press Freedom Index in last few years

• The graph shows an opposite trend to the real scenario, as the rank is degrading but the graph is moving upwards
• Difficult to find exact rank in a particular year other than the given ones(in X-axis)
• The thickness of the graph is very large which makes it difficult to read accurately

My attempt to improve the visualization :

I tried to make the visualizations as interactive as possible because of the large number of counties, which isn't possible to compare altogether in a static graph.

Graph 1 : India ranks 150th position, among all the countries (View Interactive chart)

• In the first graph I tried to visualize all the countries according to their geological location. There are total 5 geopolitical zones in the dataset which are Europe - Asie centrale, Amériques, Asie-Pacifique, Afrique, Maghreb - Moyen-Orient. I tried this to find out if there is pattern based on location, and the graph also shows there are some relation with the geography. The Europe/Asia central zone covers all the top rank countries while the Middle east countries comes in the lower rank.
• Because it's a interactive chart, one can hover over ant point to see country name, which allows accessibility to all the country's data.

Graph 2 : A drop by eight positions from 143 to 150 (View complete chart)

• This graph focuses on the difference in rank between 2022 and 2021. It is important to see how all the countries have improved or degraded from last year. The blue end is the current ranking and the orange end is the 2021 ranking
• As you can see some country's rank changed drastically and some country's remained same.
• Also the data is sorted based on the current rankings so that the viewer can navigate through the data easily

Graph 3 : Trend Comparison with other countries (View Interactive chart)

• The third visualization is to see how ranking of a country has changed over the years. A simple line chart with data points at every year which can be hovered to show more information is implemented.
• Along with it all the other country's data are also added and one can select them from the top selection menu. This feature will help the user to compare one countries rank to other countries of their choice over a period of time.

[JaanhaviSP]

Article: People in south India are far more liberal in matters of religion and nationalism

Story: People in the south of India tend to be more religious and less opposed to interreligious marriages -Even though they are equally religious compared to other parts of India, only a few consider theirs to be "One true religion." -62% of them go to places at least once a week, which is more than the share in central, eastern, western, and north-eastern parts -Only 37% of them (least among other regions) thought it was essential to stop women in their community from marrying into another religion

Stark contrast -Although southerners are more liberal in terms of dietary restrictions, only fewer of them considered a person to be not Hindu if they eat beef or a person to be not Muslim if they eat pork -Southerners have more close friends from outside their religion and caste circles compared to people from other regions -The stark contrast is that about 75% of Hindu southerners would have a Muslim as their neighbour -Religious opinions also varied highly between college-educated and those who did not attend college

Details of Data:
-Type of data used: Ratio scale -The study has been conducted across 29 states and UTs. People from all major regions, among 17 languages and all age groups excluding children, were studied.

Critique:

• The data visualization provided above is not an effective way to communicate such an important statistic.
• The yellow cells represent the prejudiced lowest, and the blue cells represent the liberal highest of the region of the country
• The two columns represent the comparison of opinions of college-educated and those who did not attend college.
• 6 essential insights are being tried to convey, but the data provided are for 14 items put together in a disorderly manner
• The stark contrast is not established properly
• The data is not encoded clearly as the insights are not conveyed

My attempt to improve it:

• I started by sorting the data for six insights. The chosen insights were those 7 points in the story
• Discarded data from the data set "It is essential to speak Hindi to be truly Indian", "It is very important to be a Hindu in order to be a truly Indian", and "Have an altar, shrine or religious symbol in their home or wearing a religious pendant" as I felt were irrelevant to the story and the stark contrast they are trying to convey
• I also separated the data into 3 sections, "The story to establish stark contrast", "the stark contrast", and "Comparison between college grads and Those who didn't graduate college", for an apparent understanding of the information

Iteration 1 for "The story to establish stark contrast":

• Created a bubble chart representing the comparison of percentages in every zone of India
• This chart displays the magnitude very evidently and is easy to compare
• It is easy to read both extend of the data

Feedback (Tested out to a few to check the effectiveness):

• People perceived that blue was cheerful extent and red was negative extend

• The magnitude was not evident

  Iteration 2: (Implementing the changes)

Graph for "the stark contrast": Choropleth map was made

Graph for "Comparison between college grads and Those who didn't graduate college":

[amitkumarram95]

Amit Kumar Ram Roll no. 216330005

Article : Assam, Arunachal and Mizoram saw the biggest increases in encroachment of forest land in past two decades

Brief Overview of the Article :

• Most northeast Indian states are ecologically sensitive zones, with more than 70% area under forest cover
• However, these states have seen a substantial decline in forest cover, due to encroachment and diversion of forest land for non-forest purposes
• More than 1,000 hectares of forest land was diverted for non-forest purposes in Arunachal Pradesh, Manipur and Assam, over the past decade
• Assam, Arunachal Pradesh and Mizoram also recorded the maximum increase in forest land encroachments over the past two decades among all Indian States.
• The article also looks at the total share of forest area in every Indian state.
• There is also a chart that shows the total forest cover (in square kilometers) in the northeastern States over the past decade, and how the trends are in terms of forest coverage are changing over time. Except for Assam, forest area has declined by varying degrees, in all the other States.
• There are also tree charts showing the total forest area diverted for non-forest purposes, as well as the increase in forest land encroachments

The article essentially aims to bring to light - how forest area has been continually decreasing over the years because of human activities, especially in northeast Indian states, which enjoy a fairly large forest area cover

The major issue I feel with the data visualizations in this article is that they do not generate any sort of emotional reactions to the issue that the article is addressing. In that regard, the color palette as well as the visualizations are 'weak', although they are practical and somewhat effective in conveying the necessary information. They also lack any sort of encoding that can help power the story.

Visualizations used in the article :

This graph shows the share of forest land in different states in India - through interactive dots. The size of the dots is constant, and clicking on a dot reveals the name of a state, and the percentage of land under forest cover in that particular state.

What is the purpose of this graph? How does it help the story that the article is trying to tell?

• I feel that the purpose is to convey how important Northeast Indian states are - in terms of forest reserves, as most of these states have a fairly high forest cover.
• Thus, establishing the above fact helps in bringing out the gravity of the situation. It tries to accentuate the fact that - the most important Indian states in terms of forest cover are losing their forests.
• The one advantage of this graph is that it smartly navigates the issue of accommodating values with a large range - by limiting the datapoints to just percentages. The geographical area ranges from a few 100 Sq. kms to more than 300,000 Sq. kms and it is a huge challenge to accommodate all these values in the same graph.

However, there are certain issues with this graph.

• This graph lacks visual effectiveness in conveying or generating any sort of emotion wrt the bleakness of the story.
• The graph does not fully utilize the space it occupies either. A large area is just devoted to small dots.
• The graph does not give any idea about the geographical area, nor about the actual forest cover. This might be interesting to note, as certain states like Arunachal Pradesh - have a large geographical area, and large forest cover cover as well.

This set of trendcharts depicts how "area under forest cover" (in Sq. km) has changed over the years, in Northeast Indian states. After much head scratching, I realized that the best way to show trends over time - is through trendlines.

I do not have much complaints about these graphs, they communicate the information perfectly.

The only small issues are :

• The scale on the y-axes are different in these different graphs. Therefore it becomes difficult to compare the change in forest cover for different states. For eg., Arunachal Pradesh's graph starts at 66,500 hectares and the y-axis labels keep increasing by 500 as you go up (66500,67000, 67500 and so on). However, for Assam - the values on the y-axis label differ by 200 hectares. Thus, even if the slope of the trendlines is similar, the actual values would be very different.
• Data is not available for the years 2012, 2013, 2014, 2016, 2018, 2020 and 2021. However, straight lines drawn through available data points might create a misunderstanding that area under forest cover has been steadily declining over all the years for which data is not available.
• Also, should more emphasis be put on a state by Arunachal Pradesh which lost more area under forest (a fairly large state in terms of geographical area), or on a state like Nagaland, which has lost a higher percentage of its forest area (a much smaller state in terms of its geographical area). Which of these deserves greater emphasis and concern?
• The choice of colors seems arbitrary. The visualizations do not convey the emotion of loss.

This is a tree graph that shows the Increase in encroachment of forest land (in hectares) between 2002 and 2022.

The focus on the article seems to be on Northeastern states.

However, I realized after observing the graph that the data is not particular to North Eastern states, which weakens the narrative.

There are several issues with the graph is terms of color, contrast and sizing, like :

• The different shades of blue used in gradation, render the smaller boxes barely legible.
• The lack of contrast between the light blue boxes, and white text - makes the text difficult to read
• Another issue with using shapes like squares, rectangles or circles - is that there is no clarity whether a viewer must compare the edge lengths or areas under the shape.
• The comparison between states with smaller data points in terms of value - becomes difficult.
• Despite the availability of space, the graph does not mention the specific data values for each state.

This graph shows the amount of Forest land diverted for other use (in hectares) in different states. It surprised me when I saw Sikkim included in this graph, and Nagaland excluded. But it seems that this graph resembles the previous, in the manner in which it does not exclusively focus on Northeastern states.

My observations and critique for this graph are similar to the ones I have mentioned for the previous graph.

Data Collection :

Link to Data : https://fsi.nic.in/forest-report-2021-details Chapter 2 Forest Cover

The data was obtained from Forest Survey Report, 2022. I then cleaned the data, created some extra columns to check values such as percentage change over time.

Explorations

There were some fundamental questions that I struggled with.

• How do you reinforce the idea in a viewer's mind that they are looking at data related to the loss of forests. What sort of colors do you use? What sort of visual encoding do you use?
• How do you show that Arunachal has lost way more forests than the rest?
• How do I accommodate datapoints that are very different in terms of scale? The scale of the data varies from 8k sq. km to 67k sq. km and in most cases, the difference is only in 100s. Geographical area (in Sq. km) also varies from a few hundreds to a few 100,000s
• Not using visualizations that may confuse users as to whether they should compare lengths or areas. For example - I wanted to show a circle shrinking in area to convey the idea of loss of forests. But then again, it wouldn't be very clear as to whether a viewer should compare the radii between circles, or the diameter or area.
• The rate at which forests are lost, or the change percentage - Nagaland has lost 8% of its forest cover in the last 10 years, however that is not reflected in the data visualization in the article. Also, what is worse? Nagaland with a relatively much smaller area losing more forests (8%); or Arunachal Pradesh with a much larger forest area, losing a smaller percentage of land?
• Using scatter plots causes a lot of overlap between circles, thus making the visualization unclear

After days of being lost and struggling with executing the ideas in my head, my final conclusion was to use a line (bar chart) to convey the information. Using a circle or a rectangle might look fancy, but will create ambiguities as to whether an area must be compared, or radius, or diameter or length.

I also did not want to genuinely do away with using trees, or shrubs to convey the idea of loss of forests.

This is a good example of the issue I was facing wrt fitting the data into a single graph. Area related datapoints for states like Tripura (a few 1000 sq. kms) and for states like Rajasthan (more than 300000 sq. kms) - are extremely different. This disparity, although not as pronounced as in the example mentioned, is still quite large between Northeastern states also.

Thus including data that varies so much, in a single graph would be extremely difficult. Which is why, much like in the original article, I decided to create separate graphs for different Northeast Indian states.

To show change in forest cover over time, I decided to go with a simple bar chart, and tried to make the bars give an impression of trees.

To show the state-wise forest cover wrt geographical area, I decided to use a Map of India and displayed the data on it. This allows users to visually see and get a sense of the geographical area of different states, and also know the percentage area covered by forests - through the intuitive color coding, or by reading the labels present on the graph.

Final Design :

[ankitgit-hub]

Ankit Anand M.Des Sr. 216330007

Article - Data | How many Indians own a fridge, AC or a washing machine: A State-wise split

The story of the data visualization

The author is trying to tell two stories here. Number one- What personal vehicles do Indians use for commuting? Number two - Do Indians have basic household appliances?

Note- For this redesign exercise, I have focused on the story of What type of personal vehicles do Indians use for commuting?

The author has taken the data for vehicle ownership from the NHFS-5 survey conducted by the government of India. This data shows the percentage of people who own/don't own personal vehicles in India segregated by State level. This survey also captures the vehicle ownership data for each state's rural and urban populations. It also captures ownership data by vehicle type. The author's intention in highlighting this data turn could be for varied reasons:

1. Vehicle ownership data highlights the extent of disposable income availability of Indians.
2. Vehicle ownership data can reveal the preferred way of commuting for a country/state.
3. The type of vehicle preferred for commuting can highlight interesting patterns about what kinds of vehicles are preferred in certain types of terrains.

The data that the author is working with captures vehicle ownership for all Indian states. It also captures data for vehicle ownership for Urban and Rural populations in each state.

Data Collection

The data for this visualisation came from the NFHS-5 survey conducted by the government of India. This data is available on the website; however, the data is not easily extractable. The data is in the form of PDF reports, and there is no option to download the data as CSV or in Excel format. Further, the data for household consumption (from which the data on vehicle ownership is taken) is not consolidated in one report but spread across 28 reports for each state. Thus I had to manually collect the data from each of these individual reports to be able to finally start working on the redesign.

Analysing the data visualization

The author uses a map visualization to show the data for vehicle ownership for all the states. I found a few pros and cons to this approach which allowed me to think about how we can represent this data in a better way.

Screengrab from The Hindu.

Pros - The visualization used is easy to understand. We can find out patterns in vehicle ownership by State.

Cons - Putting one data on one map takes a lot of screen/print real estate. The data for rural and urban ownership of vehicles is not captured. It is hard to compare ownership data between different kinds of vehicles. To find out the data for vehicle ownership for a particular state, one has to scan and remember the data from all four visualizations. Since this map is an accurate physical representation of India, smaller states do not stand out. (ex - Goa is an outlier in car ownership but does not stand out in the visualization) The visual design is very basic and is not inviting to read.

Final Design

Detail View

How this redesign works better than the original design —

• All of the data is shown in one visualization, making it easier to scan and find patterns from the data.
• The data for vehicle ownership for rural and urban populations are also captured in the same visualization.
• It is easier to find patterns in the data at a glance. For ex-, One can easily see in the redesigned visualization that the car ownership rate for all states in India is pretty low, except for Goa. This data can also be seen in the original design, but due to the physical size of Goa in the map of India, this data does not stand out well enough.
• It is easier to compare vehicle ownership data between different states. (Ex - Nagaland stands out due to its very low vehicle ownership rates across all states)
• Using an abstract shape to represent the data creates a beautiful visual pattern which invites the reader to look at the visualization.

Iterations

These are a few iterations in which I explored how to best show the data for vehicle ownership.

A few iterations on the final design.

[Bishal2288]

Bishal Goswami PhD, Batch: Spring Sem 2021 214134002

Article Link: Find here.

1. Story the author is trying to tell:

The story focusses on the diversion of forest land in the northeastern part of India to non-forest purposes in the last decade (2011-2022). It highlights the ecological sensitiveness of the northeastern states in comparison with other Indian states supported by data. To further create an urgency towards the state of environmental affairs in the region, the author also includes the data on forest land encroachments in the previous decade. Additionally, ranking (in the order of percentage of forest land diverted in the past decade) was done for the northeastern states to bring immediate and extra attention towards the situation.

Following is the list of data visualisations that supported the story: Line graph depicting the Indian state’s forest cover as percentage of its geographical area. Line graph of each northeastern state depicting the change in total forest cover (in square kilometres) over the past 1 decade (2011-2022) Tree map showing the area of forest land diverted (in hectares) in northeastern states for non-forest use under the Forest Conservation Act over the past decade Tree map showing the absolute increase in the encroachment of forest land area (in hectares) in 14 Indian states over the past 2 decades (2002-2022).

2. Critique on data and data visualisation:

Graph 1:

This is a simple line graph showing the forest cover of Indian states as percentage of its geographical area (x-axis). The northeastern states (except Sikkim) are coloured in orange and the rest in purple. Despite this being a simple line graph, one cannot find out the exact percentage of forest cover of a state unless it is hovered over the point. The graph doesn’t have limitation of space and state names and their numbers could have been accommodated without additional function of hovering. Besides, the distance of the data points from the x-axis (showing % of forest cover) could have been reduced for better readability. Also, since the plotting is done along a single line, the reality and the pivotal narrative that the northeastern states have higher percentage of forest cover and hence, ecologically sensitive, is not coming to the fore. A simple bar chart could have been more effective.

Graph 2:

In this case, the 7 northeastern states (excluding Sikkim) were selected and a line graph of each northeastern state depicts the change in total forest cover (in square kilometres) in that particular state over the past decade (2011-2022). X-axis shows the time and y-axis shows the forest cover area. The data points were from the years 2011, 2015, 2017, 2019 and 2022, for all the states. The data from that particular year is only revealed when hovered over. Clustering of line graphs was a good idea. Upon close observation, it was also realised that not all the graphs have the same interval along y-axis. As a result, all the graphs show a similar steep declining trend except that of Assam. This tells a more effective story.

Graph 3:

This is a tree map depicting the diversion of forest land for non-forest use (in hectares) in absolute terms in the 7 northeastern states (excluding Nagaland). Despite availability of enough space to write the figures alongside the state names, the diverted area figures don’t appear on the map unless hovered over. Besides, comparison and arriving at the difference of areas between similar sized states, for example, Assam and Tripura or Mizoram and Meghalaya becomes difficult.

Graph 4:

This is another tree map depicting the increase in encroachment of forest land (in hectares) in absolute terms. The map tries to show data from 14 Indian states, out of which, data from few of the states/union territories including Haryana, Uttarakhand, Andaman and Nicobar Islands and Nagaland are not seen easily. The increased area figures don’t appear on the map unless hovered over. The data from Andaman and Nicobar Islands and Nagaland don’t even appear after hovering over because of the webpage algorithm and greater degree of difference as compared to rest of the states. Besides, comparison between similar sized states becomes difficult. The choice of colour is also not wise as the gradation of this colour across the map makes it difficult to read the names of smaller chunks.

Other crucial observations:

The data sets are disparate and are not consistent with respect to timelines. One data set is from 2021, two data sets are from 2011-2022 and one data set is from 2002-2022. Some data sets exclusively highlight the northeastern states whereas the others also include data from other states. Despite these heterogeneous data sets, the author has tried to bring attention of readers to the northeastern states and its status on forest cover over time.

3. Data collation from the article:

4. Derived data from the article:

Derived data from Graph 2:

Note: The negative percentages show the decrease in forest cover from 2011 till 2022.

5. Redesign explorations:

The original data story tries to establish the following things:

• The ecologically sensitive status of northeastern states by showing the share of forest cover as compared to rest of the Indian states (Graph 1).
• Change in forest cover of the northeastern states over a decade (Graph 2).
• Diversion and encroachment of forest land in 2 decades in the northeastern states (Graph 3 and 4).

Explorations for point 1:

Explorations for point 2:

Explorations for point 3:

6. Final redesigned article with new visualisations:

7. Links to new (interactive) visualisations:

Graph 1: https://www.datawrapper.de/_/Uq48X/ Graph 2: https://www.datawrapper.de/_/3egbs/ Graph 3: https://www.datawrapper.de/_/97UKX/ Graph 4: https://www.datawrapper.de/_/BhWIr/ Graph 5: https://www.datawrapper.de/_/Jmj1D/

[IndubhusanRoy]

Name of Article: In 2021, over 1.5 lakh died in road accidents, most were young men speeding on two-wheelers

Article Link: https://www.thehindu.com/data/data-in-2021-over-15-lakh-died-in-road-accidents-most-were-young-men-speeding-on-two-wheelers/article65844935.ece?homepage=true

Name: Indubhusan Roy Roll no: 216330002

Story the author is trying to tell: The author is trying to bring to notice the fact that there was a sudden upsurge in the no of accidents post the pandemic. Though it is quite evident that the number of accidents fell during the pandemic it is surprising that the post pandemic numbers are much higher than pre pandemic times

The author provides the following graphs to support their claim

1. Line graph showing no of people injured, killed and the total accidents
2. The % share of road accident deaths in 2021 across various states
3. Share of road accident deaths by age (%)
4. Share of road accident deaths in 2021 by gender (%)
5. Cause of road accidents (%)
6. Vehicular share of road accidents (%)

Data Analysis: Graph 1

The chart shows the number of road accidents (blue), number of injured (red) and number of related deaths (yellow , right-axis) in India. As it is clearly seen that in 2021, the number of deaths surpassed the pre-pandemic levels.

However plotting the number of accident related deaths on the right axis is misleading and also creates a wrong impression.

Data Analysis: Graph 2

This map shows the State-wise number of road accident deaths recorded in 2021. Uttar Pradesh recorded the highest number of deaths followed by Tamil Nadu.

Though this data clearly shows that UP recorded the highest number of road accidents it does not provide any breakthrough information as UP is also the most populated state in India. So as an inference plotting the number of accidents to the population per state would be an interesting observation

Data Analysis: Graph 3 and 4

The following charts show the % share of road accident deaths in 2021 across various categories.

Share of deaths by age

Share of Deaths by gender

Data Analysis: Graph 5

This graph highlights the reasons for the various accidents

Again this graph though simple is not very well designed as it lacks consistency in the positions of the text and the visual style

Data Analysis: Graph 6

This graph tries to highlight which vehicles have the most accident rates and clearly two wheelers stand tall here.

However it is not clear if these vehicles have the most accidents, or deaths. It would be interesting to observe the co relation of deaths to the two wheelers and the four wheelers. Having the most number of accidents in a particular type of vehicle may not be the cause of maximum deaths

Gathering data

As per the observations marked above for the specific charts I started gathering the data for the population ranks from census of 2011 and also ranking the states in terms of the no of accidents in the year 2011 from the following sources https://ncrb.gov.in/sites/default/files/ADSI-2021/adsi2021_Chapter-1A-Traffic-Accidents.pdf

https://www.indiaonlinepages.com/population/state-wise-population-of-india.html

This was done to co relate and understand which states have higher accident rates per capita. The data set is as below

Redesign 1: Percentage share of road accident deaths in 2021 across various states

This data set was then used to create a range plot comparing the two rankings

From this plot it is quite evident that states which have the darker red plots on the left like Tamil Nadu, Odhisa and Chhattisgarh have high accident rates per capita of their population

Another iteration for the same is using a stacked bar to show the varying ranks

This further reinstates the fact that having higher no of accidents in a particular state is not correct a measure of the quantum of accidents in a given population

Redesign 2: Line graph showing no of people injured, killed and the total accidents

The line graph in the original article had issues of readability because of different quantities of the same variable on both the sides. The purpose of this graph is to highlight the sudden increase in the deaths from accidents after the pandemic. Though there is a drastic decrease in the overall no of accidents since 2014 the author wanted to highlight the sudden increase in deaths. For this the data was gathered from the original article

For this graph one of the redesign solutions is to plot all the lines in terms of the original values in a sequence. This makes the graph more readable and though the redesign does not have an element of shock it projects the data in a correct and legible way.

It is evident from the red plot line that the no of deaths have increased in an unprecedented way

The second iteration for this is plotting the data to a dot plot. Also as the no of deaths are marked in red with the years on the Y axis it provides a decent representation of the sudden increase in the no of deaths after the pandemic

Source: NCRB

[shivanginegii]

Name of Article: Data | The highs and lows in States' infant mortality rate rankings

Article Link: https://www.thehindu.com/data/data-the-highs-and-lows-in-states-infant-mortality-rate-rankings/article37181356.ece

Name: Shivangi Negi Roll no: 216330018

What is the story the author is trying to tell? In this article, the author is trying to state that the IMR infant mortality rate (IMR, deaths per 1,000 live births of children under one year of age) in India from 2014 to 2019 has reduced gradually in each state over the years in numbers. But a significant amount of reduction has been noticed in states with high IMR before 2014 compared to states with low IMR before 2014. So, Thus, the reduction of IMR in terms of absolute value cannot be used to gauge the performance of the States. Hence they tried to compare each state's IMR in 2014-2019 with individual countries' IMR in the same year range.

Low IMR - Life expectancy more High IMR- Life Expectancy is less Variables - State(location), Global Rank, IMR, Countries(location) The Census range took - 4 years

The type of data used is Ratio & Interval For that they have published 3 kinds of tables for comparison:

1. State-wise IMR (2009)

2. State-wise IMR (2014)

3. State-wise IMR (2019)

4. State-wise IMR (2009)

2. State-wise IMR (2014)

3. State-wise IMR (2019)

What should help users understand the data well:

1. Showing the evident improvement in each state over the years in terms of the shift from high IMR to significantly low IMR. This can be done by showing the changes in the before IMR value and after IMR value for each state & with the reduced number/percentage/level.
2. Although there is a significant reduction from before, the lMR is still really high. it is not mentioned in the article. Considering what the readers would want to know next could be the primary reason as to why they still persist in each state. The scope can be increased in that way as it would make them aware of the reason behind this.

What I felt did not work in the data representations: Comparing the rank of an individual state to a high-income or a low-income country will work in 2 ways:

1. States like Kerela are compared to High-income countries like the USA, they might feel positive about the numbers, and the drive to reduce the Infant Mortality rate further might not persist. They already are struggling with showing significant improvement over the states with previously high IMR.
2. States with High IMR that have made significant improvements to reach lower IMR, might not understand the concept of being compared to a Global Rank or countries that are low income.

Conceptualisation:

Design attempt : 1

Design attempt 2:

Design attempt 3: Link :https://www.figma.com/proto/ty6dfz3zpF9IZzby3u70eh/IMR?node-id=1%3A7&scaling=min-zoom&page-id=0%3A1&starting-point-node-id=1%3A2

Design attempt 4: https://www.datawra pper.de/_/IpcH5/

Final Designs: Out of all the iterations and learnings, the most direct & informative was the chloropleth map. It served the appropriate need for this article because:

1. It helped the users view the overall change across the years- 2009,2014 &2019
2. It made it easier for the readers to compare differences across all the states at a glance
3. It helped user notice changes within a state across time
4. Additional information ( created for 2019 only, for the sake of representation), like - IMR for that state, growth value between 2 years, & ranking with respect to similar countries had been represented. 5. Please note that it is an interactive map, that has been completed for 2019 only. Link: https://www.figma.com/proto/ty6dfz3zpF9IZzby3u70eh/IMR?node-id=122%3A684&scaling=min-zoom&page-id=122%3A6&starting-point-node-id=122%3A680

[dsanika]

Data | Will the roll out of AC III tier economy class bring loss-making Railways back on track? Name : Sanika Deshpande 216330011, MDes

Article Link: AC III tier was the only class that made consistent profits

Story of the Article: In the case of Indian railways, AC III tier was the only class that made consistent profits in the FY13-FY20 period. Specifically their experiment of introducing AC III tier economy class coaches has started to pay off. As per the data, adding more AC III tier economy class coaches is the right step.

The author supports this with the following data

1. Operating ratio of the Railways in the last 12 years

In this case we understand that currently railways are spending a lot of money and working on very less profits. The data in this graph makes it essential for Indian railways to take some steps to improve their profits.

2. Share of internal revenue, extra budgetary resources and gross budgetary support in the total revenue receipts of the the Railways

The data in this graph talks about the internal resources and budgetary resources which are the part of the entire expenditure of the railways. In order to make the case strong for the AC III tier to roll out, this data is not essential to tell the story.

3. Percentages regarding the passenger carried, kilometers and earnings from the passenger for all classes of service. Also including the average rate charged per passenger per km

This table consists of Important data about III tier AC. From all the classes, Earnings from it are quite higher as compared to the passengers carried. Also the charge per passenger is not the highest. This data needs to be redesigned in a better format.

4. This focuses on the uniqueness of AC III tier by showing operational losses incurred while operating various classes of service

This data is very insightful to show that only AC III tier is always in profit for Indian railways.

Conclusion: Keeping the 1st graph same as it is conveying important data Eliminating the 2nd graph as it is not essential to tell the story Creating a visualisation regarding AC III tier parts and details Completely redesigning data in 3rd and 4th pointer to emphasise more on the data

Redesigning

1. Understanding AC III tier

Through this it makes it clear that there are two types in the AC III tier. Also increasing the number of Economy class would help more as there are more berths and the cost difference is slight.

2. Comparing all class % (Passenger traveling and earnings) Iteration 1:

Iteration 2:

This will be an interactive graph. Users can view the exact percentages by hovering on the graph. This gives clarity that hardly 1% passengers travel by III AC, still it has huge profits.

3. Comparing earnings of the class with average rate charged

Through this we realize that III tier AC is not very expensive but has decent earnings.

4. Operational losses incurred while operating various classes of service

Only the AC III tier graph is in profit over the years. This would be interactive graph.

[Anngarime]

Name: Annapurna Garimella 216330010, M.Des.

Article link: India gets its 75th Grandmaster

Original design

The Story

This article was published on August 7, 2022 - soon after V.Pranav, 16 year old became country's 75th Chess Grandmaster (GM). There are 2 central claims of the article:

• number of GMs in the country has increased steadily since the early 2000s
• with time more younger people are becoming GMs The article overall uses 4 sets of data represented in various ways to support the above.

Critique

Visualisation 1: "Thambi on top" The article starts with a map showing the number of GMs each state has produced. With help of a Choropleth map, it is very effectively established that Tamil Nadu has produced the maximum number of GMs. However,

• how is this data relevant to the story of the article?
• the map conveys that southern part of the country produces GMs while the northern hasn't - this could be misleading as not having GMs does not mean that the states do not contribute to the pool of quality chess players.

Visualisation 2: "Masters on rise" This graph tries to highlight that number of GMs has increased steadily since 2000s. The graph uses cumulative data, so visually looks like number of GMs is increasing,

• Plotting a continuous line misleads the user to think there were GMs in all years.

Visualisation 3: "Younger, Faster, Stronger" This graph tries to prove that age of GMs is decreasing over years, but,

• it uses avg age of GMs - which dilutes the essence & power of the data
• also it doesn't really give a strong visual image that age is decreasing

Visualisation 4: "Number of GMs and top ratings" Tries to give a global perspective of ratings and where India stands.

• again, how is this relevant to the story?

Other remarks:

• there are years with no grandmasters - that should have been highlighted
• the increase in GMs post 2000s should be more prominent
• since we are trying to understand the historic trend of GMs, it is important to highlight the less dense decade 1988 to 2000

Redesign

What did I do?

• Combined visualisation 2&3 to convey the entire story at one glance
• Curated a clean set of data : 75 GMs and their age

IDEATION SET 1

Iteration 1 Feedback -

• How do I know how many GMs in a given year?

Iteration 2 Feedback -

• Visual clutter could be reduced
• Years skipped & corresponding number not making sense

Iteration 3 & 4 Feedback

• Who are the other players?

Reflections

• is this the best way to represent trend?
• at the first glance user is lost
• takes time to notice the density
• took time to read age

IDEATION SET 2

Iteration 1 Reflection

• better to tie all bars with x axis

Iteration 2 Reflection

• colour contrast could be improved
• font size could be increased

Iteration 3