In our amazingly fast digital world, the collection of data has become very powerful. The power in numbers to help a client operate more efficiently, to help a lawmaker make a critical vote that will impact her constituents in countless ways, or even how an individual buys a car or house – all of these decisions are shaped by data gathered, analyzed, and acted upon. I like to think of it as data provides wisdom to make more powerful decisions.
The power in data comes from the analysis itself. It just cannot be a set of numbers in an Excel spreadsheet. What do they mean? Where are they from? What do they mean in relationship to one another? What trends do you see? What trends don’t you see? How does your data compare to some baseline? These are all types of questions that must be asked by those that want to use data to further their business, government, and/or personal decision making skills. The really cool thing about data analysis is that it is an incredibly creative and exciting process.
Yes, I said data analysis is exciting. I know, I know, many of you may disagree with me, but I cannot tell you how many times I have been super excited about finding some hidden “gem” in the data that proves that a particular product should be shelved because it isn’t generating revenue, but actually costing the company money. Unearthing these gems is precisely why I love what I do as a data analyst. I think because of my background in American government and quantitative analysis, I approach my daily work like a social scientist. I think about what question needs to be asked or tested (hypothesis) and then I go about collecting the data points (variables) that I need to test that question. Once I have them, it is just then a matter of figuring out what reporting features I should use to tell the story (analysis).
The best part is that you don’t need an advanced degree in quantitative analysis or statistics to do some of these things within your own organization. You can simply start by collecting data you are interested in (or your supervisor) on a consistent, similar, and normal timeframe. If you are working with a company like GovDelivery, we actually provide you many usage figures that can easily be exported into Excel or cut and pasted right into your reports. If you have collected your own data, take a few of the variables and plot them on a line chart in Excel, over time. This trend line will give you insights into higher/lower traffic to your website, peak times when constituents are in your license office to get renewals, etc. Throw in an average from neighboring counties/cities/agencies, on the same measures, and you will see how your organization compares with those around you. Take those charts and write up a simple proposal to your supervisor regarding peak hours for the office and insert these charts into it. Provide action items at the end that can be implemented straightaway. Suddenly, you have provided wisdom for the entire office and a powerful service for the constituents that you serve.
Let me know if I can help inspire your data analysis wisdom. My door is always open and happy to help out another Govie.
Yes! Few people understand the thrill of seeing a mass of data suddenly take form into something meaningful and unexpected. 🙂
I agree, data is a very powerful tool that can give decision makers the right information at the right time to influence the right decision. The key to success is in asking the right questions and knowing the data well enough to understand it’s accuracy. A lot of data is meaningless depending on the question you are trying to answer, and many times, you may have more data than you know, it just needs to be integrated across stovepiped data “owners”. Analysis is also important. Few people have the ability to see tabular data and understand where the outliers are. Technology has come a long way and it is much easier to “see” data in geospatial environments (think Google Map) and interactive charts and dashboards. We’re doing this on the facility management side and it is a powerful tool to help build and defend a budget when there is pressure to trim funding and reallocate it to other programs. Data is exciting!
The trouble I have when approaching projects from a quantitative standpoint is the history of depending on ancedotes to make allocation decisions. It’s something I’m tackling in developing a priority-based budget as well as analyzing the need for new capital projects. It’s a slow process, but I agree that it is an exciting one. The beauty of depending more heavily on data is that chances are the data will confirm the ancedotes we hear every day. I think the challenge is selling data as a better way to allocate public funds.
I think the key is objectivity. If you have sound data collection processes that are well documented and maybe benchmarked or follow an industry-wide standard, you ensure a credible, consistent and repeatable data stream that overcomes the emotional arguements and fients to throw out your data. The questions have to align with the mission/vision of the organization and the metrics need to support those questions. On the support side of the mission, it’s often difficult to convey “non-sexy” basic infrastructure and system repairs/upgrades, however over the long run, those projects impact the mission the most.
I won’t beat the dead horse on this one (no horses were harmed in the actual writing of this post), that data accuracy has everything to do with how good your “analysis” will work out and be accepted widely. The government learned this through all of their eGov work, in so far that, sexy dashboards draw viewership..but that doesn’t mean the data is actually worth anything. Recovery.gov suffered from horrible data accuracy, still does, as it relies entirely on the data owner for entry. (no validation, no oversight, etc.)
So, I present for arguement, “Correct Data Leads to Correct and Powerful Wisdom”..where poor data only leads to poor decisions and poor analysis.
Agreed on the data input side. Garbage in means garbage out. The difficult part is determining whether the data is actually legitimate or not. If you can build it yourself, and have the understanding of what the variables are at the start, you have a much better chance of producing something amazing. If you are given data, do some examination of it to make sure it looks accurate. The smell test is often one of the first things I do with data. In response to Megan’s comment, I have had the good fortune of often proving a “long standing belief” wrong with data. Going into a manager’s office and showing that a “hunch” was wrong — amazingly powerful. As I said in the piece, I truly love what I do and how I get to dive into these questions to find some type of answer. Usually leads to more questions. Thanks for all the comments and I look forward to continuing dialogue with folks via GovLoop.
@Chris and @Dennis – Totally agree. Every manager should receive a copy of “How to Lie with Statistics” when they are promoted to management so they can defend themselves against GIGO-nlysis.
As a former proposal writer for non-profits, I LOVED data – especially when I could turn numbers into narrative and stats into stories. In some instances, it’s not the size of the stat / dataset that counts; it’s how you use it.
This may be a bit meta, but is there any data on how data is being used w/in the public sphere. @Megan mentions a common issue in budgeting, and certainly there are similarities across the board in hiring and promotions, project selection etc. There are a lot of new ways to use and look at data (let’s just assume it IS good data)… but are we using them, and if so, how?
Data analysis can be exciting! 🙂 Dennis are you a fan of Tufte?
I wouldn’t go so far as to say that there is an express-route between data and wisdom – oftimes data is an express-route to unproductive distraction – but it sure provides a nice basis for insight when you’re ready to see it and welcome it.
This particular Youtube video has received over four and a half million hits, and deservedly so. If you can’t find something exciting about working with quantitative data after seeing this, then you should probably go back to sleep.
http://www.youtube.com/watch?v=jbkSRLYSojo
+1 on Tufte
I have been known to enjoy Tufte on occasion. In terms of Mark’s comment, you definitely need to know if you are headed down a hole and how to get out of it. Data may be that distraction, but the analysis part is where I think the true magic and art are formed. I can’t describe it to folks if they have never experienced it. Finding out something totally new and remarkable, or proving an old theory completely wrong. Amazing and something I get up every day to go and have an impact doing.
If you like Tufte then you will enjoy Stephen Few – http://www.perceptualedge.com/blog/
“the analysis part is where I think the true magic and art are formed. I can’t describe it to folks if they have never experienced it. Finding out something totally new and remarkable, or proving an old theory completely wrong. Amazing and something I get up every day to go and have an impact doing.“
I work in a bland 10 x 10 cubicle, for what I consider mediocre management, and pay that is decent but not the sort of thing you’d brag to anyone about. I have my share of bonehead proof-reading to do, answering e-mail from irritable people my manager would rather not deal with (bilingually), stuffing envelopes for mailouts, meetings so dreary and devoid of purpose they would constitute a viable homicide defense. But for a certain segment of the year, I get to work with data that I helped generate. I get to roll around in it like a puppy on the grass. I get to “wonder if”, and test ideas. I get to know things that possibly no one in my entire country (and maybe even yours) has ever known before. It’s like finding an undiscovered species or people, like seeing a mountain peak no one has ever seen. I get to be the scientist I was trained to be (and I was trained to be damn good).
So yeah, Dennis, I get where you’re comin’ from, and I’m comin’ from the very same place. It’s a VERY nice place. 🙂
I think a lot of people shy away from working with the quantitative for a few reasons. First, it is generally taught by the very last people who ought to be teaching it, so it is too often poorly explained hence poorly understood and presumed to be “hard”. Second, it tends to be taught like it was catechism, and too many students don’t see it as “alive” but as a fait accompli. Third, we too rarely think of it as an epistemological tool, as a gateway to insight, to ideas, to knowledge. It is certainly not the sole pathway to all of that, but it is no less of a pathway than any other.
Megan hit the nail on the head: “I think the challenge is selling data as a better way to allocate public funds.” Knowledge is power but in the public sector, decisions are not and will never be made based solely on scientific criteria. Evidence can improve the process but it is naive to think it will substitute for politics.