Food for thought

readingsOur regular Friday diet of suggested readings for the weekend:

The best way to win an argument, as it turns out, is not to bash your opponent and tell him exactly just how much of an idiot or ignoramus he is (go figure!), but rather to ask him to explain how, exactly, he thinks his position works. Be careful, though, he may pull the same trick on you, and you may be the one who ends up changing his mind…

A long essay in The Guardian about so-called “extreme” altruism. It really makes you think. Anyone’s got a half-decent argument for why the main subject of the essay is just wrong at conducting her life the way she does?

These days it seems like everything you thought was the case isn’t, especially when it comes to food. This article may be partly in jest, but the point is important: how, exactly, is the average consumer supposed to settle on a healthy diet, given that nobody seems to know what that looks like?

Finally! A half decent argument for why people don’t recognize Clark Kent as Superman…

Should we stop worry and learn to love A.I.? This essay makes a lot of good points, but I think it still (unbelievably) misunderstands the Chinese Room argument.


Categories: Plato's Suggestions

38 replies

  1. Massimo, yes on both those specifics. Wikipedia (sorry, Dan Kaufman) has a good entry on effect size, which per this conversation, specifically notes its value in **social and medical research.** I think confidence intervals “track” more to p-values than effect size, so if you asked me for one focus, it would be that.


  2. One needs both effect size and confidence intervals (because if the effect size is large, but the confidence interval is huge…). The latter are definitely better (and less arbitrary) than p-values.


  3. Off the “effect size” link I saw this about the MAGIC guidelines. Pretty interesting.

    Liked by 1 person

  4. Comparing effect size with the p-values is a little bit “apples and oranges”. Each are useful in different ways when used appropriately.


  5. Massimo,
    do comments get held up in moderation because they contain links to external sites?


  6. Robin,

    no it isn’t actually. p-values are supposed to tell you whether it is worth paying attention to a given effect size, once one considers the likelihood of getting such effect by chance. For a number of reasons that I won’t go into here (but are well discussed in the literature), looking at effect size and confidence interval is a much better way to go. p-values may be misleading, and their threshold is arbitrary (and many people, including professionals, insist in confusing statistical significance with scientific significance).


    yes, comments with links are held in moderation to minimize spam. I usually get to them within 24 hours though.


  7. Massimo, right. I’ve heard others mention your POV before. The arbitrary threshold was my point on p-values, but this is likely better addressed by other statistics. And, that MAGIC link I think parallels in part your ideas on statistical vs scientific significance.

    Liked by 1 person

  8. Thanks for the reply Massimo, and sorry for the delay. You told me:

    I don’t think anyone is denying it, but we do reject that this person’s obviously selfish behavior should be honored by the rest of society.

    Yes I certainly do agree, and this is given that the rest of society will be a distinctly different subject. If you agree that personal good is ultimately based upon personal qualia, then you must also agree that social good is ultimately based upon social qualia. Furthermore apparently social good may be just as repugnant as individual good can be, though I shan’t elaborate.

    The point is that here we are agreeing upon the nature of what’s ultimately good/bad for the conscious entity, and yet you’ll not find such a formal understanding in our mental and behavioral sciences. Not only do they remain agnostic on this subject, but presume that science need not determine this in order to effectively do their work. I, conversely, believe that instrumental good/bad must become the premise which founds these fields. Observe for example that the modern psychiatrist’s therapy sessions were so ineffective, that they’ve essentially been replaced by medication. Furthermore the modern cognitive scientist seems no closer to providing us with a functional model of the conscious mind.

    This all ties in nicely with the next post. I’ve been anticipating your coming book since you mentioned it in April when we first met…


%d bloggers like this: