Lady Time at the Bus Stop

They often refer to Time as a female character. Indeed, she plays a significant role in the theatric drama we call Data Quality. A while back, I was in a hurry to get some updated retirement account balances and had to reset my password in order to get into my account. I provided my email account and requested the password reset, expecting an email to shortly follow with instructions and a hyperlink to change my password. The problem was that I didn’t get the email, with the link to reset the password, until the next day (long after I needed the information). Clearly this is an example of a functional failure of the Financial Services provider’s website (but also a data quality issue because I didn’t receive the reset password link (data) in a timely manner. The Conformed Dimensions define this Underlying concept of Timeliness as:

Time Expectation for Availability- The measure of time between when data is expected versus made available.

In a prior blog about the relationship between user interface design and data quality I touched on how important data Representation is, and this time I’d like to draw the connection between functional design and data quality. As shown in the example above, more and more applications are data driven- meaning that depending on the data provided, or from the application’s database, you see and interact with the application differently. That means that if the data isn’t correct (note I didn’t say accurate which has special meaning in data quality), you may get unwanted results.

Another example of timeliness also comes to mind regarding the medical field. A family relative was in the hospital to receive an infusion of an expensive drug. The nurse provided a shot of premedication but forgot to chart the fact. Another nurse came on shift and would have given the same premedication again had I not reminded her that my relative had already received it. She said there was no record of that- and proceeded to call the prior nurse to validate it and get an approximate time of injection. Arguably, the first nurse was going to input the data (perhaps after her lunch break) so the data was just untimely, but clearly patient outcomes were at best improved by not receiving another shot and at worst experiencing a life threatening over dose, though unlikely I was assured).

Are you ever in a hurry? How about when you ride public transportation? That has always been my issue with public transportation here in the USA. When I lived in Japan I enjoyed the consistency of arrival/departure times for trains and generally for buses, so when I saw this LADOT Commuter Express sign, I was encouraged that there is hope for public transportation. Note that this illustration gives us examples of both Timeliness versus Currency. As the case with complex systems today, the bus's onboard computer may be logging its location in real-time (sub-second observations) but when that data is provided via the ladotbus.com website is a measure of Timeliness for end-users. In short, currency measures the speed of recording the real-world concept but Timeliness measures its effective use taking into account the availability of the data for consumption/application.

Note that the data can be:

• Current (reflect the real-world) but not be timely. For example, an hour after the bus is scheduled to arrive (the app updates to say “ALL ROUTES CANCELLED FOR THE DAY”). Assuming that indeed the bus wasn’t running that day and only the message that they had been cancelled was late, we’d say the info is still current- reflecting the ongoing fact that the busses aren’t running- but clearly of much less value, given it’s so late. The rider could have made other plans, rather than wait for an hour before giving up and making other plans.
• Timely (arrive at the frequency expected) but not be current. For example, the public transportation app may provide updates of the current location of the bus every 5 minutes (near real-time), but waiting riders may be disappointed if they appear to have arrived before the bus, only to find out that in reality it has already passed the bus stop.

Timeliness and Currency are often least appreciated until needed. Send me (dan[at]DQMatters{.}com) your stories about Timeliness and Currency so that we can strengthen our collective effort to articulate data quality effectively.