Usability vs. Usefulness

Usable and/or Useful Data?

Imagine that a cop uses a radar gun to catch speeding cars, but that the radar gun is completely unpredictable. This faulty radar gun does one of the following, with no predictability:

  1. determines the car’s correct speed
  2. provides no reading at all, or
  3. shows an incorrect speed

The unpredictability of this imaginary radar gun would make the data it produces useless. This brings us to the crux of this activity: not all data is useful or usable.

Usable Data

What makes data usable?

Data is usable if you can use it, regardless of whether that data is informative or useful.

Another aspect of usability is whether the means to process or analyze the data currently exist. Consider the case of the SETI Institute, which has been collecting data from radio telescopes for years and years, but has only been able to analyze roughly 2% of it (despite the vast amounts of computation leveraged to do so). This data may not be considered usable in its current form, because it’s simply too big and messy.

What’s interesting about usable data is that in many cases, usable data need not be completely predictable or accurate as long as it falls within a range of possibility. An example of this type of data would be GPS data for phones. With the latest smartphones, GPS can be accurate up to about 10 meters (at its best), which isn’t perfect, but is definitely usable...and useful.

Useful Data

What makes data useful?

Data is useful if somebody would want to use it, in essence making it valuable for some purpose or another (not necessarily financially). The usefulness of data is also directly related to what somebody can do with it.

Imagine a light sensor that gives the exact wavelength of light (color) emanating from one square foot of the ocean.

This sensor provides an exact, real-time reading, but who cares? Can you make predictions, describe some process, or solve a problem with this data, or is it simply a bit of miscellany that’s not of value to anyone? Not surprisingly, the usefulness of data is open to interpretation and context.

Imagine someone has a whole array of these sensors spread across the entire Pacific Ocean. In this case, the amount of color in a grid square can be compared with the data collected from other sensors. Is this useful? That may depend on whom you ask. To a criminal lawyer, car mechanic, or accountant, it may be useless, but to a marine biologist, it may be extremely valuable and useful.

Imagine again that one of these sensors has collected data regularly for 30 years, and over that time, the color has shifted from green to blue in certain geographic areas.

What might the reason be for the ocean in this place to obtain a greenish tint? What inference might be postulated with this data? The mere existence of these questions lends credence to the likelihood that this data could be be useful.

When determining the usefulness of data, perspective matters.

Common misconception: If data are being collected, they are being utilized.
  • Data are being collected, because they can possibly be utilized. However, just because data exist does not mean that they are being used or utilized.

Instructions

Think about the following scenarios and how usable or useful the data described can be. Challenge yourself to view the questions from different perspectives, and be prepared to discuss afterward!

Survey

1) Define usable in your own words.

Answer:

2) Define useful in your own words.

Answer:

3) How usable would the receipts from every McDonalds transaction in a day be?

  1. Extremely usable
  2. Somewhat usable
  3. Not usable at all
Answer:

4) How usable would your family’s medical records be?

  1. Extremely usable
  2. Somewhat usable
  3. Not usable at all
Answer:

5) How useful would 30 random social security numbers be?

  1. Extremely useful
  2. Somewhat useful
  3. Not useful at all
Answer:

6) How useful would every grade that you’ve ever received be?

  1. Extremely useful!
  2. Somewhat useful
  3. Not useful at all
Answer:

7) How useful would the results from hooking a robot up to a polygraph machine be?

  1. Extremely useful!
  2. Somewhat useful
  3. Not useful at all
Answer:

8) Which of these data sets would be the most usable? Why?

  • the results from hooking a robot up to a polygraph machine
  • every grade that you’ve ever received
  • 30 random social security numbers
  • your family’s medical records
  • receipts from every McDonalds transaction in a day
Answer:

9) Which of these data sets would be the most useful? Why?

  • the results from hooking a robot up to a polygraph machine
  • every grade that you’ve ever received
  • 30 random social security numbers
  • your family’s medical records
  • receipts from every McDonalds transaction in a day
Answer: