MCAT psych section bank 53 Help

[holdmystethoscope]

---

Members don't see this ad.

I am not sure if I a allowed to post screenshot or type up the question here due to copyright issue. If you have access to the aamc section bank psych section please help me out 🙂

Question 53, why the answer wouldn't be B. Why can't we say the hunger and sleep deprivation are correlated! Thank you so much.

 

[flyingeagle]

I got this wrong at first too, but it is because they used an experimental design with an independent variable (lack of sleep). Thus, their conclusion can be a lot stronger than simply "they are correlated". In fact, the experimental design allows you to conclude there is a causal relationship between the two.

 

[holdmystethoscope]

I got this wrong at first too, but it is because they used an experimental design with an independent variable (lack of sleep). Thus, their conclusion can be a lot stronger than simply "they are correlated". In fact, the experimental design allows you to conclude there is a causal relationship between the two.

Thank you so much. I never thought of this way. When would you think two things are correlated vs they have casual relationship? I know if correlation has a value r=1 that means they are positively related. but in terms of correlation vs casual relationship how would the wording be different in the question. Thanks again

 

[flyingeagle]

Thank you so much. I never thought of this way. When would you think two things are correlated vs they have casual relationship? I know if correlation has a value r=1 that means they are positively related. but in terms of correlation vs casual relationship how would the wording be different in the question. Thanks again

You need to be able to identify it from the type of experiment they performed. The reason experimentation is so good is because it allows you to reach causation. So, if there is a manipulated independent variable that leads to different effects on a dependent variable, it is a causal relationship. Other data usually leads to correlations (e.g. survey data, retrospective study). These types of studies don't manipulate anything and are just observing whether two variables are related. Or, they assign groups based on predetermined groupings (e.g. sex, race) and look for correlations that way. These are not experimental designs because you can't manipulate someone's sex or race.

 

[holdmystethoscope]

You need to be able to identify it from the type of experiment they performed. The reason experimentation is so good is because it allows you to reach causation. So, if there is a manipulated independent variable that leads to different effects on a dependent variable, it is a causal relationship. Other data usually leads to correlations (e.g. survey data, retrospective study). These types of studies don't manipulate anything and are just observing whether two variables are related. Or, they assign groups based on predetermined groupings (e.g. sex, race) and look for correlations that way. These are not experimental designs because you can't manipulate someone's sex or race.

Thank you that was extremely helpful. I am little shaky on the experiments but still trying to learn as much as I can.

 

[Wolfpack93]

You need to be able to identify it from the type of experiment they performed. The reason experimentation is so good is because it allows you to reach causation. So, if there is a manipulated independent variable that leads to different effects on a dependent variable, it is a causal relationship. Other data usually leads to correlations (e.g. survey data, retrospective study). These types of studies don't manipulate anything and are just observing whether two variables are related. Or, they assign groups based on predetermined groupings (e.g. sex, race) and look for correlations that way. These are not experimental designs because you can't manipulate someone's sex or race.

I missed this question for the exact same reason, I'm so used to "correlation is not causation". So it's safe to say that if it's an experiment and the IV shows significant change due to change in the DV then that can be described as causation?

 

[flyingeagle]

I missed this question for the exact same reason, I'm so used to "correlation is not causation". So it's safe to say that if it's an experiment and the IV shows significant change due to change in the DV then that can be described as causation?

Yeah I'm used to that phrase too, and correlation is the more conservative answer so it is probably right more of the time. But yep! If you have a randomized, independent variable that causes different effects on a dependent variable, then you could conclude causation. Just remember not to be confused with quasi-independent variables like race, sex, SES. These are not randomly assigned variables, so you can't conclude causation for these types of studies.