Visual.ly – How does it work?

On 12 March 2012 infographics startup Visual.ly launched (with some fanfare in the visualisation community) a new suite of infographic visualisation tools. Visual.ly have had some large investment rounds, with over $2m invested so far, based on their plans to democratise visualisation. I took a look around to see what is on offer.

Visual.ly are creating a community that submits, reviews and rates data visualisations and infographics, provides learning materials and articles for content creators and links up data visualisation experts with businesses and advertisers.

The new visualisation tools allow users to use pre-packaged visualisations to present their data. Currently the examples available are statistically very simple with some interesting styling. They allow basic analysis of Twitter and Facebook statistics.

Here is a promotional video that explains the concept.

The Visual.ly website describes the new tools:

BIG NEWS. THE WAIT IS OVER. INFOGRAPHIC CREATION TOOLS ARE HERE

Visualizations are powerful, but they’ve required time and hard work to create -- until now. You no longer need expensive software, extensive design skills, or number crunching ability. We’re building a tool that will allow everyone to quickly and easily create professional quality designs with their own data. And when you’re ready to show your work to the world, publish it on your Visual.ly profile, your own personal showcase.

Lee Sherman (Chief content officer at Visual.ly) said they’re trying to marry the ease of use that exemplifies Mint with the greater world of data out there. And while right now that means just making some instant graphics using Twitter and Facebook data, that will soon mean access to various APIs and publicly available data sets, as well as letting users upload their own data and even mash up data sources. Ultimately, Sherman said, users will be able to move away from pre-packaged infographics and actually edit the fields themselves. (gigaom.com)

Where are they headed?

This all sounds mighty exciting! After hunting for information I am left to conclude that there is not enough out there at the moment to see where Visual.ly is headed. This has left me to ponder….

It sounds like the big idea is to pre-package visualisations and get people to plug their data into them. This would do for infographics what a calculator did for simple maths. Instead of working it out yourself, you can type in some numbers and the answer is created.

To create more complex statistical models for visualisations that allow users to plug in their own data would produce some interesting and complicated programming challenges. If visual.ly head down the route of making complex charts, images and graphics that can be used to produce enlightening views of data, this would be a commendable endeavour with great utility. I also love the idea of datasets being made available on their website for people to interrogate and produce graphics with. That sounds like fun as well as potentially opening up huge datasets for the populous to dissect and analyse.

I will be interested to see if the site is filled with visually interesting but statistically childish charts. Content of low statistical quality will be much easier to create. It will also be much easier for users to apply to simple datasets as not much numerical understanding is required. Instant access to high impact charts with a low statistical utility could be used in many a presentation to liven things up, but may not be much use in providing new perspectives using data. Fancy graphics lacking in substance could even obscure interesting analysis if emphasis shifts from statistical rigour to pretty presentation.

The big question for me is this: 

Will the site improve the use of data in infographics and visualisation? Or will the site shift the emphasis of infographics towards design rather than statistical content? 

One thing is for sure, it will be interesting to see how it develops.




Simon Conway is a co-founder of Condaal and an evangelist for Lean Information. He can be reached on simon.conway@condaal.com

Hack rape stats shocker!!! (Journalistic misuse of statistics – Part 1)

From time to time I stumble across an article that so thoroughly abuses statistics, I feel compelled to write something down. I spotted an excellent case in point in the Metro on 28^th February.

http://www.metro.co.uk/news/891565-serial-rapists-stay-free-because-of-failures-by-police-and-prosecutors

Here’s a summary:

The title proclaims: “Serial rapists ‘stay free because of failures by police and prosecutors’.” The title sets the scene. After reading this title I am expecting an expose on the failures of law in the UK and horror stories of serial rapists stalking the streets. Pretty gripping stuff!

The subtitle embellishes this with “Serial rapists may be escaping justice because of a raft of failures by police and prosecutors, a damning report claims.” Currently we are still very much on the theme of expose and serial rapist terror.

We enter the body of the text and statistics start being bandied about:

“Their attacks go unrecorded because as many as three in ten reported rapes are written off as if no crime ever took place, the investigation reveals.” This sentence states that as many as 3 in 10 rapes are assigned as ‘no crime’ committed. The use of the emotive words ‘as many as’ and ‘as if no crime ever took place’ gives the impression that this is a scandal. I have no idea if this is a scandal or not. I have nothing to compare this statistic to. The language informs me that a scandal has taken place rather than the meaningless statistic.

“Figures showed almost 12 per cent of rape cases in 2010 ended with police ruling no crime had been committed compared with just three per cent of GBH cases.” A few sentences later we are presented the national statistic of less than 12% of rape cases assigned as ‘no crime’ committed. This is much lower than the 30% rate that was quoted above. Why the drop? This is cleared up in the following sentence:

“In Kent, the figure was 30 per cent although it was just 2.4 per cent in Gloucestershire.” It appears that rather than using the larger, more accurate data set at the start of the article the reporter cherry picked the biggest number to sensationalize. This latest comparison begs the question ‘Why is there such a difference in

‘no crime’ judgements between Gloucestershire and Kent?’ The question remains unanswered.

“More than one in ten of these decisions were wrong, according to an analysis of at least 100 such ‘no crimes’, the report says.” Wait one second! This statistic shows 1 in 10 of the rape cases assigned as ‘No crime’ is incorrect. So of the 12% of rape cases assigned as ‘No crime’ nationally, 1.2% are incorrectly assigned. We have dropped from 30% in the first sentence of the article to 1.2%. This is disguised through confusing words. It definitely did not jump out at me. I read the text several times to decipher the conflicting stories being told by the statistics and the text.

“Interpol record checks on rapes abroad ‘were not regularly conducted’, even though they can identify a pattern of offending.” This sentence is the justification for the serial rapists. The incidence of serial rape by foreigners in the UK would have been interesting line of enquiry. As it stands the article gives no clue as to the size of this problem. The title persuades us towards the seriousness of the issue. The data gives us too little information to refute the titles proclamation.

Journalism as weak as this is undermines the story that it tells. Without clear facts about a story how can a serious dialogue develop? A quick scan of Wikipedia and related articles on the internet gives a sense that there are serious issues with rape convictions in the UK. A home office study in 2005 states some facts we know about rape:

Since the age of 16, 7 per cent of women had suffered a serious sexual assault at least once in their lifetime (5% had been raped).

The paper reports ‘an estimated annual incidence rate of 47,000 adult female victims of rape.’

“Home Office figures show an on-going decline in the conviction rate for reported rape cases, putting it at an all-time low of 5.6 per cent in 2002.”

A number of studies have found high rates of ‘no criming’

These are serious issues. Unfortunately Metro and Nick Herbert haven’t provided the facts, clarity or dialogue to support the cause.

Opening post

Welcome to the Condaal blog.

Bringing you our take on lean information.

We want the world to understand this vital concept and why it is important to business, the world and life itself.

We will have links to lean information and information that is out of shape, reviews of software that we love or hate and anything else we find interesting.

Enjoy!