Howdy, Stranger!

It looks like you're new here. If you want to get involved, click one of these buttons!

Phenomenon hypothesis & Correlation causation intensive

MarkmarkMarkmark Alum Member

Hi all the purpose of this post is twofold: to teach everything I know about phenomenon hypothesis (PH) and correlation causation (CC) argument types, as well as to be a knowledge check where other people can correct me and make sure my understanding is solid. I'll start with PH, then go through CC, and then I'll show how both argument types are very similar.

In PH arguments, we have something that happens in the real world, then we offer an explanation of why that thing is the way it is. For instance, I see a bunch of seals barking. Then I see fishermen riding their boats in the harbor. I say, "It must be that seals bark whenever they see fishermen." My phenomenon is "seals bark" and my hypothesis is "they bark when they see fishermen." In lawgic, the SC would be "See fishermen" -> and the NC would be "Bark."

There are a lot of ways to strengthen PH and CC arguments and I'll explain them here: 1. A->B, 2. block B->A, 3. block C->A&B, 4. block "no relation," 5. block bad chronology, and 6. show "good consequences." 7. No cause no effect

  1. A->B really just means "If see fishermen -> bark." This works with PH and CC arguments. How can I show that seals really do bark when they see fishermen? Show more data. a trend of more data. I don't want to see 1 more case of seals barking when they see fishermen, I want to see a lengthy trend of seals barking many times over a long period of time. I have seen at least 2-3 times where the LSAC will use a trap answer where the "strengthening" answer choice just throws in 1 more example of the hypothesis working. "You say seals bark when they see fishermen, well Joe saw a seal barking when fishermen were present." I want to see "over the last 5 years, there's an 85% chance that seals will bark whenever they see fishermen." I don't want a single corroborating example (although this does strengthen the hypothesis very, very, very slightly), I want to see a trend.

  2. Block B->A. Let's say my argument is "When the sun shines, then my trees grow." To show that B actually causes A is a little weird in this case, but it would go like this: "My trees growing actually cause the sun to shine." If the latter case were true, then my argument that "sun shine -> trees grow" would be ruined! The causality would be flipped the other way around. The B->A style works really well for CC arguments where I'm trying to show that A is causing B; to show B->A, or block B->A can weaken / strengthen the argument.
    For example - "When the sun shines, -> trees grow." To strengthen this argument I can block B->A. "It's also not the case that trees growing causes the sun to shine." I'm eliminating the possibility that my causality isn't flipped. To go back to seals, I would block the case that barking (NC) actually is the explanation for the seals to somehow be seeing fishermen (SC). "It's not the case that barking allows the seals to see fishermen."

  3. Block C->A&B = block an alternate explanation.
    What if it's the case that shrimp actually cause the seals to bark and the fishermen to appear? In that case my phenomenon hypothesis argument would be ruined. It's not the case that seeing fishermen causes seals to bark. It's something else.
    I want to block this alternate explanation: "It's not the case that shrimp cause fishermen to appear and that shrimp cause seals to bark."
    In a correlation causation argument, let's say "hearing about earthquakes in the news causes people to dream about earthquakes." But what if everyone was watching a movie about earthquakes, and this movie caused the dreams? We would want to strengthen our argument by blocking an answer choice that says "A recently released movie about earthquakes is known to cause people to dream about earthquakes." We can eliminate the possibility of an alternate explanation, and this strengthens our argument that actually hearing about earthquakes in the news caused dreams.

  4. No relation / 5. bad chronology
    Bad chronology goes hand in hand with "No relation" so I'll group them here. "If I study -> get 180." What if I see an independent study that says "studying has been shown to have no effect on your test results."? That would show "no relation."
    Likewise, what if I said "Bob studied then he got a 180. Therefore, studying gives you a 180." Then I say "Bob started studying AFTER he got a 180." This shows bad chronology - the effect actually occurred before the alleged cause! Block this to strengthen.

  5. Good consequences
    If the phenomenon hypothesis argument is true, I want to show good consequences. If my hypothesis is true, what would happen? Let's say my hypothesis is "If seals see fishermen -> then they bark." Good consequences AC would say "Fishermen have increased in the bay 500% in the last month. Since then, sales of earplugs have increased 1000%." Fishermen are in the bay a lot more, which means seals are barking, and people don't want to listen to that so they buy earplugs.
    Another example is "The city is increasing the speed limit by 30mph. Therefore there will be more car wrecks." A good consequences answer choice would say "There have been more speeding tickets since the speed limit increased."
    If we take our hypothesis to be true, then make a reasonable assumption of what could happen if the hypothesis were true, we get a "good consequences" answer choice that strengthens the argument.

  6. No cause no effect
    Let's take the argument "It's sunny. Therefore my ice cream will melt." If S -> M.
    If I show that there's no sun, and indeed my ice cream is not melting, this is good! I strengthen my argument.
    However, if I show "Cause with no effect," or "Effect with no cause," these WEAKEN the argument. It's sunny, but my ice cream isn't melting. Or "my ice cream is melting, but it's not sunny!" These weaken.


Sign In or Register to comment.