A1-KesavaMallela
From CS294-10 Visualization Fa07
Contents |
[edit] The Good One
The hidden wealth of poor, The Economist, Nov 5 2005, pg.2. Originally, from a World Bank report by Thorsten Beck, Asli Demirguc-Kunt and Soledad Martinez.
- Explaination
The graph is used to show the strong correlation between lack of access to financial services and low incomes. It plots various countries ranging from Bangladesh to Singapore to illustrate the point.
- Deconstruction
The data set for the graph includes the following set of data marked in the graph: (i) Estimated % of households that have bank accounts (ii) GDP per person PPP in '000 USD from 2003. The graph plots about 20 countries on it, but the regression line (in red) claims to consider data from 54 countries. The graph uses linear scales on both axes.
- Critique
The visualization does an excellent job of showing correlation at the lower ends. The regression line shown in red neatly joins many countries that do poorly at both access to financial services as well as PPP of an individual. If I were to modify anything in the graph, it would be the following couple of things:
- The graph does NOT explain why only 54 countries were considered. My preliminary guess is the graph plots countries from Asia, Latin America and Eastern Europe. Spain is an exception though.
- The graph also does not explain the anamolies that are too far way from the regression line. e.g. Mexico seems to be in the same league as Chile in terms of PPP, but Chile has a much accessible banking system than Mexico.
[edit] Bad One
Image Source: India Today
- Explaination
The graphic appears in a leading South Asian magazine to highlight problems being faced by new capacity additions to an already overloaded power generation grid in India. The graphic points to several new additions of significant capacity and calls out the problems being faced by each of them.
The audience of the magazine are diversified. People looking at the graphic are either looking at capacity additions to the part of the country they live in OR for a big picture of problems being faced in general.
- Deconstruction
The data set used here are (i) the geographical location of the new additions (ii) power generation capacity of each addition (iii) Qualitative information like the name of the power company operating the addition as well as the nature of the problem being dealt with. Data types (ii) and (iii) are displayed together. Both of them are depicted as qualitative information in the original graphic.
- Critique
The original graphic uses the geographical location of the additions to show which part of the country they fall in. It mixes the other key quantitative data available, the power generation capacity, with qualitative information. This does not help the reader navigate the problems in the decreasing order of power generation capacity, thus making it harder to gain a big picture. Further qualitative information is too fine grainly color coded, forcing the reader to refer to the legend every single time, making it even harder to gain a big picture. Also, the icons in the legend are not intuitive enough. Too much of emphasis on the poor power bulb metaphor which eats away lot of space.
[edit] Re-design
India map taken from from WikiProject Indian Maps
- Critique
The redesign tries to use the power generation capacity of each addition to help readily differentiate between additions with significantly different generation capacities. Three different sizes grid-tower icons have been used to mark ranges of generation capacities. Thus the reader can quickly skim through problems being faced by additions with large generation capacities with out actually looking at the every single call out. The possibilities of color coding of callouts have been reduced to make it easier for the reader to identify the broad issue. More specific details about a capacity can be included in its callout.
[edit] Mediocre One
| Athlete | Sport | Statistic | Standard deviations | Probability against (1/x) |
|---|---|---|---|---|
| Bradman | Cricket | Batting average | 4.4 | 184,000 |
| Pelé | Football (Soccer) | Goals per game | 3.7 | 9,300 |
| Ty Cobb | Baseball | Batting average | 3.6 | 6,300 |
| Jack Nicklaus | Golf | Major titles | 3.5 | 4,300 |
| Michael Jordan | Basketball | Points per game | 3.4 | 3,000 |
| Joe Montana | American Football | 3.1 | 1,000 |
A table to make a strong point that Sir Donald Bradman is the greatest player in history of all ball games. Using the above criterion, it is estimated that someone of Bradman's calibre appears once in 184,000 batsmen, compared with once in 9,300 football players for the next most highly rated person, Pelé.
Listing the second best player in each sport would have made the table even more persuasive as it would give different audience better benchmarks to compare with.
Source: Wisden and Wikipedia
