Random Rants 92 - Not Enough Snerk

[Synobun]

My nose looks like Rudolph the Red-Nosed Reindeer because of a clogged pore that's grown overnight into a raging infection. It's very swollen and painful now. I drained some, but what I got out doesn't seem to match the size, so there's likely more to come over the next days. In the meantime, I've hired out my services to the North Pole.

 

[Takhisis]

The rants thread has almost been pushed to the second page. What happened? Was there a major outbreak of positivity on here?

Four days without posting, T-boy.

You could also interpret it as all ranters being too upset to post here, if it makes you (un)happy.

 

[Synobun]

France got horrifically robbed in Eurovision this year.

 

[Kyriakos]

My nose looks like Rudolph the Red-Nosed Reindeer because of a clogged pore that's grown overnight into a raging infection. It's very swollen and painful now. I drained some, but what I got out doesn't seem to match the size, so there's likely more to come over the next days. In the meantime, I've hired out my services to the North Pole.

Wow.
I hope it passes soon

 

[Snerk]

 

[The_J]

I drained some

...you need to pay a lot of attention there. Apparently all the veins running through the nose also run through the brain. Messing around in this area can easily give you an infection even higher up, which you really don't want to have.

 

[Samson]

...you need to pay a lot of attention there. Apparently all the veins running through the nose also run through the brain. Messing around in this area can easily give you an infection even higher up, which you really don't want to have.

Is this true? In that, is there a risk of getting a bacterial infection of your brain lancing a pustule on your nose that is not present doing one on say your bottom?

 

[Birdjaguar]

IIRC the nerves involved in smelling go directly into the brain. Not sure about blood vessels though.

 

[r16]

taught 35 years ago at school . Do not squeeze or whatever the zit things on your head , ı think , above a line that matches your jaw .

 

[Kaitzilla]

Is this true? In that, is there a risk of getting a bacterial infection of your brain lancing a pustule on your nose that is not present doing one on say your bottom?

I've heard rumors that really bad teeth health quadruples (x4) heart disease risk.

Crap getting into the bloodstream continuously doesn't have to be from a needle.

Ya, no nose picking either.
Delicate blood vessels there.

I've never heard of popping pimples on face being dangerous, but maybe?

 

[Birdjaguar]

I've heard rumors that really bad teeth health quadruples (x4) heart disease risk.

Crap getting into the bloodstream doesn't have to be from a needle.

There is a teeth/heart connection. After my heart valve surgery, I have to keep my dentist informed. Infection in one's mouth can move to the heart.

 

[Samson]

I've heard rumors that really bad teeth health quadruples (x4) heart disease risk.

This is very true, but it is via the system circulation.

IIRC the nerves involved in smelling go directly into the brain. Not sure about blood vessels though.

The nerves do, the blood vessels certainly do not in the same way.

 

[Synobun]

...you need to pay a lot of attention there. Apparently all the veins running through the nose also run through the brain. Messing around in this area can easily give you an infection even higher up, which you really don't want to have.

How do you think I have my opinions about shows and movies?

 

[The_J]

That would be funnier if we were talking about brain worms .

Is this true? In that, is there a risk of getting a bacterial infection of your brain lancing a pustule on your nose that is not present doing one on say your bottom?

I checked it, and while I got the exact biology wrong, the principle is there that it's dangerous, https://en.wikipedia.org/wiki/Danger_triangle_of_the_face .

Lancing would be a tad bit different. If you're messing around there yourself, you normally ...er... apply pressure, which might push infectious fluids in the surrounding tissue. If the problematic area is properly lanced, this should not be the case.

 

[r16]

this needle thing ı first heard of when ı was lkke 40 or 45 . Tech is a good thing .

 

[Samson]

There was talk in one thread of the Dunning-Kruger effect, so I though I would look it up. This is what the graphs that google shows look like:

Spoiler Google Dunning-Kruger effect images :

This is what the actual Dunning-Kruger paper showed:

Spoiler Actual Dunning-Kruger effect images :

These are not the same, yeah? Also, why would someone do a t test of ranks? Is that not totally losing the whole point of parametric tests?

 

[The_J]

I guess there's a meta-Dunning-Kruger effect too .

The google pictures make it more clear what is meant, although they're inaccurate. If you took them, you'd say that people in the middle are underestimating their average skills. What is not what the D-K says. But if you showed the normal population the real D-K figures, the first thing you'd notice is that everyone rates themself roughly similar, which doesn't send a brutally obvious message.
EDIT: Looking at the graphs, I guess you could make a ratio out of the 2 values in the D-K graphs and end up with the google graphs, no?

For your question regarding the stats test, the paper isn't really clear about that (e.g. can't find what actual test was used), but I would guess that the testing itself was done on the actual values, but the display/grouping is done in the quartiles.

 

[Samson]

I guess there's a meta-Dunning-Kruger effect too .

The google pictures make it more clear what is meant, although they're inaccurate. If you took them, you'd say that people in the middle are underestimating their average skills. What is not what the D-K says. But if you showed the normal population the real D-K figures, the first thing you'd notice is that everyone rates themself roughly similar, which doesn't send a brutally obvious message.
EDIT: Looking at the graphs, I guess you could make a ratio out of the 2 values in the D-K graphs and end up with the google graphs, no?

If you squint you could say there is a drop from 2nd to 3rd quartile in figures 2 and 3, but there is no significance test and with N's of around 50 I think there is no chance is is significant. Also the graphs show an increase in confidence in the first few percentiles, with is nowhere to be seen.

For your question regarding the stats test, the paper isn't really clear about that (e.g. can't find what actual test was used), but I would guess that the testing itself was done on the actual values, but the display/grouping is done in the quartiles.

The Method 1 stats are described so:

These estimates were not only higher than the ranking they actually achieved, paired t(15) = 10.33, p < .0001, but were also marginally higher than a ranking of "average" (i.e., the 50th percentile), one-sample t(15) = 1.96, p < .07.
​

I agree it is unclear, and it should be described in the methods, but is they mean they tested the underlying values they have phrased it very oddly. I am not really sure that "score given to how funny a joke is" can really be expected to be normally distributed a priori anyway.

 

[The_J]

TBH that's also an extremely odd choice to measure "skill". Tests 2+3 are a lot more suitable.
And yes, can't read much out of hte methods.

For the first part, while I think the sample size should be okay-ish (Mann-Whitney test says you'd need 20-25 per group, so half should at least give an indication), the rest is indeed not given.

 

[Samson]

TBH that's also an extremely odd choice to measure "skill". Tests 2+3 are a lot more suitable.
And yes, can't read much out of hte methods.

For the first part, while I think the sample size should be okay-ish (Mann-Whitney test says you'd need 20-25 per group, so half should at least give an indication), the rest is indeed not given.

I agree with your 1st paragraph, but to clarify your second, are you saying you have done a power calculation, and given a true difference of the observed drop from 2nd to third quartiles (about 5%?) a 20-25 sample size will give you a significant probability (we use 80% for our power calculations) of a p value below 5%? If you have really done that I guess I will believe you, but my intuition says that does not sound right.