Click. click. click – Blink. blink. blink
It’s the sound of clicking away that all statistical data users – expert or amateurs, single or regular – know and experience. And nowhere does it plague data users more than when faced with difficulties in finding the data to do a job and achieve an outcome.
Our knowledge of user jobs comes from our user research, which include:
- being aware of available data across the GSS
- understanding and manipulating statistical data at different geography levels
- comparing data from different topic areas to generate better insight
- understanding the difference between datasets of same titles or topics published by different departments
- understanding the comparability between datasets
Whether users or user producers, these are jobs irrespective of our expertise levels or tools needed for data manipulation. If only it was easy for users to complete these with minimal clicks and effort.
Don’t you just hate it when:
- you click for so long but still don’t find the data you need
- your awareness of available datasets are limited and you don’t know which are comparable
- there is no single authoritative place to find the data
- you can’t seem to make sense of the technical language
- different producers use different definitions or terminologies
- there is a change in the data but nothing to explain what those changes are
- you spend a lot of time cleaning and formatting the data
- you find useful datasets but the similarities between these datasets leave you confused
Based on the user research we conducted, different user types embark on different user journeys. For example, a heavy user knows the data, where to find it and how to use it. What they doesn’t know is what other datasets are available and comparable to increase the value and reach of the data they regularly use.
For a non-heavy user, the challenges are knowing where to find the data, identifying the right data and understanding the data due to the dataset titles, descriptions and technical language used. Without good titles and descriptions, users tend to abandon their data journeys or, in a similarly challenging scenario, they end up downloading lots of datasets until they find the “one”. This journey can easily become a cyclical never-ending odyssey – one that I call the vicious find-review-abandon cycle!
Imagine doing this with 10 or 20 different datasets and the:
- user’s experience and thoughts during this, including frustration
- time, effort and energy wasted
- administrative overheads for producers to resolve user queries
- impact on the environment – it counts, you know!
In my best-ever rallying call: “Better statistics, better decisions!”
It is time to break the vicious find-review-abandon cycle”. It is time to improve the Government Statistical Service (GSS) statistical data provision. It is time to make the provision of statistical data more user driven and increase the reach and value of data. Yay! Cue: now is the time to run with me.
Here comes the cavalry!
The GSS Data Project is aiming to pioneer a better way for users to connect with statistical data. We want to transform how data is published within the GSS to improve discoverability, interoperability and usability, increasing the impact and reach of United Kingdom (UK) statistical data, driven by the Better Statistics, Better Decision: Strategy for UK statistics: 2015 to 2020. The long-term aim for the proof of concept is to develop a product with supporting services to standardise and harmonise the provision of Tidy Data to enable easy human and machine interaction. We are not promising we will meet all user needs, but we will cater for a good range of user jobs.
Our challenges so far
As with all transformation projects we’ve had our fair share of challenges. However, one major challenge has been the lack of existing user research across the GSS community.
With little or no user feedback available initially from departments, we had to rely on assumptions and our knowledge of the statistical landscape. We had to adopt an hypothesis-driven approach to enable quick prototyping.
Our hypothesis is:
We believe that transforming the GSS statistical data into a standards-based web friendly format will result in a more effective use of the data, increasing its reach and value. We will know we have succeeded when the agreed alpha scope of data is transformed into a machine readable format, available and can be easily discovered with a stripped down user interface by the end of June 2018.
What we know:
- it has been difficult to effectively use statistical data due to its presentation
- as an organisation (GSS) we needed to improve the dissemination of statistical data
- linking the currently silo statistical datasets can increase reach and impact
- improving the dissemination will increase data value and enable effective reuse
- there are cost savings to be realised across the GSS community
- users have got used to using the data in its current state, and any changes will impact both users and producers of statistical data
What we are learning:
- the pain points along the user journeys
- how users want to interact and be presented with official statistics and associated metadata
- the measured impact of the current provision on users
Where are we with the Proof of Concept (PoC)?
Using this hypothesis-driven approach, we have a product blueprint and a backlog of user needs to keep iterating on. This backlog has been supplemented by the user research we have been doing within the project, and we are quickly building up our understanding of the user and continuously recruiting research participants to help shape our thinking and validate any assumptions.
To achieve our definition of success for Alpha, we have had to adopt a 70 to 30 approach to developing the PoC, with more time spent on understanding and transforming the data and less time developing User Interface features. Simply, getting the data right is very important to meeting user needs, and in turn, having a good understanding of user needs is helping us get the data right.
Get in touch, we are recruiting
Calling all statistical data lovers, users and producers visualization and innovation fashionistas. We need your help to make data better.
Please contact us if you answer “yes” to any of these questions:
- Do you share the same passion we have to find a better way to disseminate statistical data?
- Do you reuse statistical data and find it difficult to find and understand it?
- Would you like to be involved in making the availability of UK statistics better?
- Do you want to reduce the number of effort required to find the data you need?