At a recent summit held by Mashable and the UN there were several issues that were brought up regarding the collection and analysis of data from certain groups. This would seem fairly obvious for many, for instance the poorest countries in the world are going to have considerably poorer data collection capabilities than the richest ones, but it goes beyond this.
Even in some of the richest countries in the world there are fundamental flaws in the way that data is collected for many groups purely from the way it is collected and the way it is input.
In most developed nations the most common way to collect large amounts of data is through censuses. The basis of these is to get the perspective of all sections of society and help gain an understanding of the problems facing people in different geographies and circumstances.
The problem with this is that these censuses take place once every 10 years, and if we use the UK as example, the last was conducted in 2011, meaning that the data in it is four years and almost an entire parliament out of date. Since this time the acceptance of many fringe groups, such as homosexual and transgender people, has increased, but the data about them isn’t accurate.
Equally, in many households there will be a single person who fills out of the census, meaning that the others in the same residence may not have their concerns fully aired. This could be because they do not want to reveal their concerns to those closest to them, i.e if they have not told parents about their sexuality or if it is embarrassing to tell housemates.
Another aspect that was discussed at this same meeting was that there was not enough being done with the data once it was collected in order to make it actionable.
Those who ultimately make the decisions based on this information need to actively understand what it means, something that is not always easy. Making sense of billions of data points and then being able to communicate this to somebody who is likely to be relatively data illiterate (compared to what would be necessary to understand this in it’s native form) is difficult, yet the best data scientists do not normally work for governments.
Therefore platforms to make this process simple are necessary, but are currently widely unavailable or unaffordable for under pressure government departments across the world. This is a situation that needs to be addressed as collecting data is essentially useless unless a government has the skills to analyze and then use it.
Essentially, data is what causes change to be made, be this data in terms of being told something by an individual or receiving the results of a survey taken by millions. How this data is collected and interpreted dictates what action will be taken, and how accurate this interpretation is comes from how well this data is analyzed.
If we get the wrong data, it will mean that analysis cannot be accurate then the interpretation and action with will be wrong. If we get the right data, but the analysis conducted poorly, then we will equally end up with poor actions. We need to look at how data is being collected in every country and international organization to allow decision makers to make accurate and informed choices moving forward. According to several findings from the summit, it may not currently be good enough.