jonwiseman

2 года назадdiscuss

Errata: this is probably not right, the pairwise trials aren't independent.

2 года назадdiscuss

So, yeah, but an interesting chunk of results: is the inter-rater agreement ACTUALLY random?

(Not sure if this is formatted correctly, feel free to comment if you'd do these stats differently)

    from scipy.stats import binomtest
    # stats: {experience_level: (rate, num_in_group)}
    stats = {
        "lvl1": (20.5, 156),
        "lvl2": (22.2, 66),
        "lvl3": (23.3, 50),
        "lvl4": (21.2, 39),
        "lvl5": (20.8, 12),
        "lvl6": (28.3, 5),
    }
    
    for lvl, (rate, num) in stats.items():
        num_tests = num * (num - 1)
    
        res = binomtest(int(rate * num_tests * 0.01), num_tests, 0.2, alternative="greater")
    
        # with multiple comparison correction with bonferroni
        print(f"{lvl}: p-value = {res.pvalue:.4f} {'*' if res.pvalue < 0.05 / 6 else ''}")
    
    print('* indicates p-value is < 0.05 after bonferroni correction')

    """
    lvl1: p-value = 0.0276 
    lvl2: p-value = 0.0002 *
    lvl3: p-value = 0.0000 *
    lvl4: p-value = 0.1337 
    lvl5: p-value = 0.4826 
    lvl6: p-value = 0.3704 
    * indicates p-value is < 0.05 after bonferroni correction
    """

So maybe there's an internal consistency in how these people are trained, and maybe it's not completely dependent on skill level. This is assuming I read the https://www.clearerthinking.org/post/can-astrologers-use-ast... part correctly.

jonwiseman

3 года назадdiscuss

Thanks OP for the post. I'm curious as to how this handles/you plan to think about "conditional" relations between nodes. E.g., for the Ginger <-> Nausea connection, is there a way to describe this link itself as being enriched in a subset of patients, such as just pregnant people? It looks like you derive association scores via text scraping then turning some text analysis into postive vs negative associations -- I'd note that lots of studies include inclusion and exclusion criteria by which this could be explored.

jonwiseman

3 года назадdiscuss

I might have misunderstood them, but I tried to argue that I was indeed converting units. Namely, 1 inch + 1 cm = 1.254 in (cm->in implicitly converted).

Also, defining a Hz as 2pi/s seems to be mixing units and quantities, like a Hz is 2 [pi/s] (where [] stores units), where pi is a unit equal to 6.28 [unitless]. Intead, defining a Hz as 2pi/s is to me changing the definition of Hz, unless it's like 2pi / 2pi [1 / s]. >You could also define one rad/s as 1/(2pi s) I was, specifically 1/2pi [1/s].

But as to your point that things like Hz are semi-meaningless unless attached to physical concepts, I'll agree with that (using a potentially different definition of semi-meaninglessness.)

My original statement that 1 Hz + 1 radian/second (i.e. 1 [1/s] + 1/2pi [1/s]) is both physical (semi-meaningless, meaning somewhat meaningful and fully meaningful when we flexibly attach it to a large range of physical concepts) and non nonsense (again, semi-meaningless until we attach to a physical concept) I think still hold.

The key to me is that we can attach physical concepts to units and measures flexibly (but not arbitarily). Hz is indeed applicable to non-sin functions, like square waves, but then we're all of the sudden talking about "the Hz of the nth harmonic of the square wave" (which has loads of relationship to angles).

jonwiseman

3 года назадdiscuss

That sum to me is both physical and not nonsense, and I think your interpretation is wrong. 1 radian/second != 1 s^-1, but instead ≈ 0.159155 Hz. This is from interpreting a radian as a ratio, in which "doing" one radian in one second is only going over 1/2pi of a whole in a second. So 1 Hz + 1 radian/second, physically and sensibly to me, means ≈ 1.159155 Hz.

comments