econ & spl & env → os

econ & spl & -env → os

econ & spl & -env → os

econ → os

correlation:

doesn't show for some reason (leftwards arrow, equals sign, 'co', equals sign, rightwards arrow, superficially looking like a weird biconditional indicating the tentative nature of the relationship, once causation is established, if established, the above suggested causation sign can be used).

My theory for the last scenario and general teaching suggestions

What if wealth is correlated with cheese consumption per capita, and with an elderly population, and an elderly population is correlated with bedsheet entanglement death?

If so then:

==ca==> cheese consumption per capita

Wealth of nation

==ca==> older population ==ca==> death by entanglement in bedsheets

I suppose because there is no =ca=> cheese =ca=> old =ca=> bedsheet death 'straight' link, there is no causation. There is correlation but both cheese and death are separately caused by the same thing 'wealth', but that does not mean one causes the other, just that both have the same cause, which if you go back far enough is true of everything. I think the above 'fork' presentation shows this well visually.

Suggested notations:

causation: =ca=>

correlation:

#feedback

good

Also (previous lesson):

A → B

A —m→ C

B ←s→ C

In the previous lesson it was said:

"Logically, any sufficient condition can be kicked up into the domain as long as it's satisfied or if you anticipate that it will be satisfied."

Option 2: /purpose and resident → prohibit

Option 3: Domain: /purpose

Rule: resident → prohibit

It makes more sense to me given the above to use RB as the exception. Obviously we're talking about residents of The Beresford. So I personally would rather go with:

Domain: resident

Rule: /purpose → prohibit

1. P: Thomas admits that he went from his apartment to the library this morning,

2. P: and there is no way for him to get from his apartment to the library without going past the Municipal Building.

3. IC: Therefore, he must have been in the vicinity of the burning Municipal Building.

4. P: No one could have been anywhere in the vicinity [Municipal Building fire] and fail to notice it.

5. C: Thomas must have seen the fire that destroyed the Municipal Building.

I agree with you that 4 is a MINOR premise. It gives support to 5. but it recieves support from nothing. Its claim is perception and that is simply supported by real world knowledge concerning our senses and the nature of massive fires. 2 just tells us he was in the area, which on its own does not ground perception. As you said it is simply an interjection, a minor premise, unsupported in the argument (accepted as true) and doing nothing other than giving support (together with 3) to 5. The only major premise/intermediate/sub-conclusion is 3 because it gives support to 5 and receives support from 1 and 2.

IMHO (obviously). Hope that helps.

Correlation allows for nonconforming data points. They do not rely on nor advertise a 100% no exceptions hit-rate. That would not even necessarily mean perfect causation cause reversal, third cause causing both etc. So some are not etc. does not touch the correlative statement. Smokers have a 45% chance of lung cancer, non-smokers a 15% chance. Both have better chance of not getting it but smoking 3x more, hence correlation. Saying but hey I know a guy in the 55% smoker no cancer group or in the 15% non-smoker cancer group does not touch the fact that smoking does correlate with an 200% increased relative chance of lung cancer.

Plane crash 99% chance of death, but aha 1% chance of no death so no correlation - nonsense.

Don't know if helpful but Ellen Cassidy says when stuck with Weaken question ask yourself: 'Does this make the conclusion less likely to be true?'

Maybe try waking up earlier. Puts a whole different perspective on the day.