For those of you who follow Stephen Few, many of you will be familiar with his opinions on pie charts. I whole heartedly agree with him, and I’d like to thank Stephen for illuminating many of the foundation principles of data visualisation and dashboard design at his course I recently attended in Melbourne. Essentially what Stephen says about pie charts is that we have great difficulty comparing areas, and we are very good at comparing lengths, so when making a part to whole comparison, use a bar chart - or if you are a little more sophisticated a Pareto chart. He demonstrates this very well here;
Upon leaving that course, I thought about this quite a bit in light of all the other ways he had taught us to simplify any data visualisation and strip it down to its purest form. The minimal information that it was conveying. The more I thought about this the more I thought that there was one piece of information in the pie chart that was not easily presented in the bar chart - the fact that you are looking at a part-to-whole relationship. When the maximum value on the X-axis in this example is 50%, it isn’t immediately apparent that adding up A to F would result in 100% - it is however in the pie chart. So I thought long and hard about how to resolve this piece of missing information. Behold, the Pareto chart which Stephen showed me, which essentially does this.
Again, there is a trade off here. By including the line to indicate the contribution to the whole we can quite easily read a cumulative total and see individual contributions. I guess you do lose a little bit of clarity for the uninitiated, as to what this is exactly - given that there is a line and bars such that it isn’t immediately apparent that they are related pieces of data. You also have a scale problem which makes it slightly harder to compare lengths when you necessarily need to make the axis go to 100%. Despite these drawbacks, I’m a big fan of the Pareto chart.
Finally, I thought that, given the fact that we now read everything on a screen, be it computer, tablet or e-book, would it be possible to get all the benefits of the bar graph, and communicate the part-to-whole relationship using animation. The stacked bar graph clearly shows a part to whole relationship, but makes it incredible difficult to compare length. If you could animate it to disassemble into a normal bar graph, you would have the best of both worlds. Eg Animate this;
to this
This is really only applicable in a Powerpoint presentation as it wouldn’t make any sense in a dashboard or report, and you do however lose what Pareto offers in assessing a cumulative total. Despite this, I think it may be an effective intermediary between Pareto and the bar chart that retains the one useful piece of information in a pie-chart.
]]>Phone (New Zealand Support) : 09 415 2447
Mike, I've been ranting about this for a long time, and at last, I believe I've found a teacher that will listen.
Obviously, there are plenty of good reasons to become a teacher - I used to be an "educational entertainer" and thoroughly enjoyed taking part in inspiring children to learn. However, as Ken Robinson says in his TED talk, () and as you have pointed out here, the system is broken.
Even the NAPLAN system which is there to add transparency to the situation caused a great backlash from the teaching institution because of the fear of this transparency, and what it would mean to the scrutiny of their jobs. In many respects, I agree with the teachers on this, despite my passion for open data, because of the methodology used for doing comparison and benchmarking with NAPLAN (although it is a good start).
The main reason I sympathise with teachers on this is that there is increased scrutiny, little feedback, and no upside. If we are going to go to all of the effort of measuring the performance of our students (and in the process the performance of our teachers) then they should be rewarded according to that performance.
I've said this for a long time and heard the groans that there is no really valid way of doing this measurement because there are so many other factors that could influence the outcome of a students education. However, there is a methodology and it is outlined exceptionally well here
http://www.latimes.com/news/local/la-me-teachers-value-20100815,0,258862,full.story
]]>If we used standardised testing to measure teachers, and rewarded them for the respective change that they made in their students education then quite possibly we would find ourselves in a situation where outcomes were better, and teachers actively sought out opportunities to teach in difficult schools where the opportunity to make a bigger difference was possible- and hence reap a great reward, both through intrinsic and extrinsic motivation.
We´re told democracy creates stability and raises living standards. So why not introduce it on a worldwide scale?
What would it take to apply the principle of “one person – one vote” to everyone on the planet? We have held 26 international field tests in every region of the planet to answer this question.
This investigation involved close collaboration with Electoral Commissions, Government Ministries, Educational Institutions and Civil Society Organizations and we would like to thank them all for their invaluable contributions.
I’m a big democracy supporter, and a big believer in big ideas. So much so, that I’ve given the global democracy problem a lot of thought.
The problem with what is contained on the World Vote Now pages, and promotional material (besides it being impossible to find much relevant information or comment) is that there is not any indication of the proposal for how the voting or the democracy would work, or what power the legislative body would have. The whole concept just doesn’t seem to be rationally thought out at all any further than “lets see if we could get electronic voting machines to everybody”.
Considering the the booming mobile phone growth in the third world is doing a pretty good job of getting “electronic voting machines” to everybody, I’m not even sure it is worth considering the current edge cases of people with no connectivity to the Internet. If the power of one vote actually became a significant agent of change, and they drastically wanted that change, then I think their priorities may shift to try and reach a point where connectivity was possible, long enough for them to vote. There is significant evidence to show that the economic choices that people in the 2/3rds world make puts mobile connectivity quite high on their short list of essentials including food and shelter.
Most of this information is moot without first considering;
- What are people voting for?
- Who are people voting for?
- What authority would the body have?
- What responsibility would that body have?
- What is the organisational structure of the body?
- What is the mechanism of change of the body?
- What power would the body have, and where would this power come from?
- Is the current mechanism of democratic delegation (currently used anywhere in the world - the US is a horrendous example) the most appropriate for global governance?
- How is the system audited, funded, and maintained?
- How do you manage unique identification of 1 person per vote to prevent gaming?
From looking at the website, not a single one of these questions is answered, but surely one could hope for more from the documentary.In any case, I think I’ve come up with a relatively good solution that I’m going to hand over to the web, or some wealthy venture capitalist, in the hope for a better world.Proposed SolutionOne person, one vote sounds good in theory, but how about one person, 6 billion votes (or the amount of the entire world population) - voting is optional, and voting lines are open all day everyday.The great benefit of this is that it makes the distribution of power far more liquid and more frictionless than it is currently. You might think that 6 billion votes is preposterous, but it would mean that the more people that one person votes for, the more diluted his power becomes. If you think about 1 person who voted for a 6 billion of the worlds population then everybody would tie with 1 vote and nobody would win the race, hence there would be vote dilution.The same applies when this is scaled to 6 billion.The benefit of having voting available at all times mean that you get a direct reflection of the consensus of the Governed at all times. If they don’t like the John Howard sending our military to fight a war on flimsy evidence of WMD’s then unvoting is a very simple proposition to deprive elected leaders of the necessary political capital required to wage a war and maintain consent of the Governed.This gives a great fluidity to Government, which some may argue would reduce stability. I disagree with this assessment, it would just mean that for stability you would require a clear lead to maintain stability - which, given the nature of the Internet I believe would be easily achievable by successful leaders.Another benefit of using this methodology is that you are no longer relegated to laborious processes of even participating in the race to represent. All people would be eligible to leader, all people could attain votes - thereby leaving the job of governing not only to politicians, but to all citizens. Not just for people with political ambition, but with sincere civic desires. It is also able to be audited by the user by watching vote counts move up and down in real time as a result of their actions.Granted, the financial rewards in this type of system maybe less, but the value of the social capital would be considerable.Implementation initially may be a little difficult - reaching critical mass, managing authentication etc, however, it may be that we are almost there. A vote/unvote mechanism is almost available via Facebook, which currently has a very well authenticated user base of 500 million (1/12 of the global population). The main drawback is that their like/unlike API doesn’t currently support a callback on the “unvote” - but it will come soon.Given this, and the slew of Twitter/ Facebook/ Email addresses and social pages, a large group of people would already be able to use this voting system immediately.The question remains, what power would this body of elected people have?Well, this is yet to be determined, but it could be agreed that all of the people that opt into the system of voting, work behind the goals of those who garner enough democratic power above a certain threshold. This would have immediate impact of current political arrangements. More ambitiously, with enough people opting in, the elected leaders would have a moral authority to discuss with media and international leaders about solutions and problems that currently does not exist. Also, since they would command no standing army, and be globally distributed, they would not pose any threat in violent terms to the status quo, increasing their legitimacy, and potentially posing a threat in political terms to the status quo in many nations.After some growth via Facebook, other mechanisms for more solid verification and more easily distributed controls and transparency would be instituted.I’d love to hear feedback on this fairly raw blog post about an idea that has been percolating for a very very long time.Cheers,Dame
]]>This image Lasoo from the edge of the Internet reminded me a lot of this beautiful image of earth from Saturn.
You can see the live data here http://nmap.org/favicon/.
Coles, I love you, but I think you have failed to appreciate the two-way communication that makes social media special. Maybe “Chat with us” would be more appropriate…
Responsibility and authority
Many people struggle at work because they want more authority.
It turns out you can get a lot done if you just take more responsibility instead. It’s often offered, rarely taken.
(And you can get even more done if you give away credit, relentlessly).
]]>Charles Web Debugging Proxy Application and packet sniffer, is ridiculously hard to find on the web due to its name but is also exceptionally handy for mobile debugging. I’m posting it here to help its pagerank.
]]>I just read a common mantra, “People start with search” – and it is true they do start searching with “search” – it is somewhat of a tautology. Before they start searching they are socialising, and if Google isn’t careful, they could very well be searching at Facebook or Twitter for the very things Google famously will find for us.
]]>
site_242_rand_1279066999_sbs_fifa_world_cup_telecast_schedule.pdf (116 KB)
The most important document you will read all year.