Source: Nestler Analytics

Category: Bad Math, Data Analysis, Digital Media, Research

Please use the comments to demonstrate your own ignorance, unfamiliarity with empirical data and lack of respect for scientific knowledge. Be sure to create straw men and argue against things I have neither said nor implied. If you could repeat previously discredited memes or steer the conversation into irrelevant, off topic discussions, it would be appreciated. Lastly, kindly forgo all civility in your discourse . . . you are, after all, anonymous.

13 Responses to “The Plural of Anecdote Is Not Data”

  1. rd says:

    You have no future in politics.

  2. george lomost says:

    Wrong. Unless the anecdote is pure gibberish, and therefore not an anecdote in the true meaning of the word, it most certainly is a datum. You can choose to ignore this and make rational excuses for this but it most certainly is a datum. And if enough anecdotes point to the same thing you’d be foolish to ignore them.

    Maybe you believe that only numbers are meaningful as the picture implies?

    • VennData says:

      george,

      “… if enough anecdotes point to the same thing you’d be foolish to ignore them…”

      You’re getting into a whale of a parameter-seeting issue if you believe your quote.

      http://en.wikipedia.org/wiki/Inductive_reasoning

      Inductive reasoning is a tricky way to make money in the markets. Set an asset allocation, some stocks, bonds, real estate, then re balance it once a year.

      P.S. This is the funniest headline of this young year.

  3. Anonymous Jones says:

    I’m sorry, but I have such a pet peeve about this little “aphorism.”

    It’s not that I don’t get what it’s supposed to mean. I do. I completely do. It’s just that it is such a bastardization of language and truth to get there that it’s more trouble than it’s worth. What a terrible idea it would be to start defining the broad term “data” as exclusive of those things you don’t wish were there. That’s how one gets to confirmation bias.

    The main point is, a group of anecdotes *absolutely*, *positively*, *undeniably*, *incontrovertibly* qualifies as “data”. There can be no reasonable dispute about this. It is a necessary conclusion from the structure of the English language and plain meaning of the word.

    Yes, it is true that a group of anecdotes *rarely* qualifies as *compelling* data.

    And yes, a group of anecdotes will rarely fit the qualifications for the data *set* one is trying to investigate.

    This may seem a small point, but I think not. The problem with having an aphorism be literally *untrue on its face* is that it will be extremely prone to misinterpretation.

    Many times, the world only provides one with small sample sizes. In that case, all one has are anecdotes. And it is better to base your decision-making on those anecdotes than *no information at all*. This is just a longtime, mostly unavoidable heuristic we must adopt to navigate the world with things we neither have the pristine data sets (or time) to get right.

    Of course it is better to use methodically gathered data sets than haphazardly gathered ones. That does not, however, change the plain meaning of the word “data.”

    • krice2001 says:

      “Many times, the world only provides one with small sample sizes. In that case, all one has are anecdotes. And it is better to base your decision-making on those anecdotes than *no information at all*.”

      Yikes, I can’t agree with that, Jones. Making decisions based on the fact, for instance, that everyone you know is in general agreement on something can be far worse than making a decision with no information. We see politicians all the time make statements about “everyone they know” but often everyone I know does not agree with that. Who knows who is right. Small sample sizes are inherently problematic which is why they’re typically avoided for providing a basis on which to make significant decisions. Not that we don’t. It’s human nature to take those convenient short cuts because of limitations of our brains. But I could never agree with your statement.

    • alonzo says:

      With all due respect, I don’t think you “get it.”
      Yes, methodically collected anscdotes may turn out to be data but that’s the point, isn’t it – mere anecdote, no matter how many, by itself, can’t be data.
      Think of this as literature not mathamatics.

  4. hue says:

    how many anecdotes does it take to screw in a light bulb?

  5. Ponchovilla says:

    Barry:
    How can you deny that a string of anecdotes does not qualify as “data”?
    For example in the hyper inflationist “theory” that gold should be at least $1800/oz by noon tomorrow based on “fundamentals” as evidence by these anecdotal data:
    I paid $1.00 more for chicken wings this week at the grocery store
    the box of cereal I bought was 10c more per box and the box size was reduced from 14oz to 12 oz
    The Federal Reserve has been “printing money” like crazy. All that QE has to go somewhere and it should be going to gold and the gold mining stocks

    Or: The “arctic vortex” has given us some of the coldest temperatures on record therefore global warming is a hoax.

    ITMT Australia has broken the highest recorded temperatures in December 2013 breaking the highest recorded temperatures of Dec 2012
    Get the point mate?

    • Because they are a biased collection of specific stories told from a specific vantage point — lacking in objectivity, filled with selective perception, created from a biased sample set, etc.

      • Ponchovilla says:

        Pefect. That’s exactly what politicians do ! I only wanted you to smile about my comment, but the reply is just too damn clear for the anecdotalists. They will never “getit” so forgetaboutit.

  6. samjoy says:

    A very useful and valid point that has special relevance in investing (ex: hyper inflation fears). A few points though:

    1) Don’t get washed away in a flash flood because the weather data called for sunny skies. It seems to me that relying on data all of the time could lead to fragility rather than anti-fragility. Data does have limitations. Making money often involves betting before the data turns.

    2) Just because data suggests something doesn’t always mean we should act on it. Data might suggest Wal-Mart should fire 75% of its workers and replace them with cyborgs in order to increase profits. In reality that would leave millions jobless and probably cause a boycott. Or imagine if data suggested we could maximize efficiency by doing something morally wrong. I see this becoming more relevant in political discussions as well.

  7. sdpost5 says:

    I just had a sudden vision of this conversation taking a turn so abstract and philosophical that it will leave us all reeling and dazed. Because it is very difficult to separate out real data from data that only appears to be significant. The market incorporates, to various degrees, everything that happens in the universe — other galaxies, etc. albeit indirectly. Freud, Einstein, and others discussed whether it’s possible for human beings to really objectively observe ANYTHING. Thus we’re left being unable to perfectly separate out what is significant and legitimate from what is, for lack of a better term, noise. OK, I have to stop here before I get myself into trouble and wake up with a hangover in Ensenada or some place.

  8. kaleberg says:

    Anecdotes are useful for figuring out what people are thinking. The ones that proliferate are the ones that reflect the zeitgeist.

    Data are useful for finding out what is going on.