The polls leading up to the election were wrong. What do you do it you can’t or don’t get accurate information?
“Media Face Backlash for Getting It Wrong,” The Wall Street Journal, November 10, 2016 B1. Media acknowledges “oversights” in run-up to Trump’s election victory.
Does this reduce the value (or cost) of future election polls? Why did so many polls get it so wrong, and so consistently? Is there a fundamental disconnect? Was there some other bias in the polling methods or analysis?
What is an exit poll worth today, versus last week? What can the polls that did get it right charge the next time? Should future polls have much larger margins of error?
Filed under Accuracy, Analytics, Business Case, Collect, Communicate, Data quality, Definition, Information, Management, New Implications, Reliance, Use, Value
Analytics are one way through massive collections of information. But do they taint the results?
“Algorithms Aren’t Biased, Coders May Be,” The Wall Street Journal, October 15, 2016 A2. Coders may include hidden or unconscious biases in the metrics they select, which affect the reliability of the “decisions” algorithms make for you.
Can you rely on a black box too much? Do you understand the devices you use and how they work? Does somebody? Can you provide oversight of a process you don’t understand?
Two articles in today’s post, both related to hospitals.
The first is, “For Hospitals, a Lot of Information Goes a Long Way,” The Wall Street Journal, September 26, 2016 R3. The more non-medical information hospitals have on you, the better they can predict outcomes. Marriage of Big Med and Big Data. Or at least they’re dating.
The second is,“Wrong-Patient Errors Called Common,” The Wall Street Journal, September 26, 2016 A2. Getting mundane bits of information wrong, like patient’s name, diet restrictions, and medications, can be deadly.
What’s the value of better information governance in the medical setting?
Filed under Accuracy, Controls, Corporation, Data quality, Duty, Duty of Care, Employees, Governance, Information, Internal controls, Reliance, Risk, Value
One of the challenges of information governance in a corporation is dealing with the stupifying amount of information that gets accessed, generated, and received by the business and its employees and agents. What if the challenge were to do that for an entire city?
“Singapore Is Taking the ‘Smart City’ to a Whole New Level,” The Wall Street Journal, April 25, 2016 R4. Singapore is a city-state that has more than 5.5 million residents. It uses sensors and cameras to track residents, cars (and when and where they are used), construction, cleanliness of public spaces, smoking, littering, and wind-flow patterns. Some information to be made available on a online platform.
What do you collect? What do you use? How do you protect it?
What can you learn from this?
Filed under Access, Analytics, Business Case, Communications, Culture, Duty of Care, Governance, Investor relations, New Implications, Oversight, Privacy, Reliance, Security, Use
Admiral Michael Rogers has an interesting view on information security. He wonders when the bad guys will transition from just stealing information to changing the information. The changes will cause even more disruption.
“It’s Only a Matter of Time,” The Wall Street Journal, October 27, 2015 R7.
What if someone chose to alter, rather than delete, your information? How do you respond?
Would the agency in charge of gathering data on the budget deficit “massage” the numbers to make the deficit look worse?
“Greeks Investigate Statistics Chief Over Deficit Figure,” Wall Street Journal, March 23, 2015 A8. A prosecutor filed criminal charges alleging falsification of data (they don’t have 18 USC §1519 in Greece) against the head of the statistics agency. Was the deficit 4% of GDP in 2009 or 15%? More than a rounding error.
What happens if government numbers are a political football? Who’s watching this?
Filed under Business Case, Collect, Compliance, Compliance, Compliance, Compliance Verification, Data quality, Duty of Care, Governance, Management, Oversight, Oversight, Reliance, Risk, Use
One form of information is metrics, or measures. But what do the metrics really measure, and are they being spun a bit?
“TV’s New Metrics,” Wall Street Journal, October 17, 2014 D1. Lots of data about the new TV shows, and how many and what people are watching them. Which one has the most viewers? Between ages 18 and 49? Is that different from the biggest audience? Which is the funniest? Who has the highest score? The most Twitter followers?
Common to the superlatives is a word with an “st” in it, e.g., most, biggest, funniest, highest. Borrowing from Richard Levick‘s work on crisis management, on how communications capture eyeball share. What information do you collect to which you can apply an “st” word to make your claim stand out? And how do you recognize the limitations on the “st” claim, when you are the reader?
How is all that information being used?