That is the question. I have found that JY will sometimes ignore logical indicator language in a stimulus (always, the only, unless, etc...) for no reason whatsoever and not really explain why he chose to ignore those indicators. Instead, he'll use simple argument analysis. Did I miss a lesson? I'd love to know what criteria people have come up with to know when it's more effective to avoid using Lawgic. JY, you're more than welcome to answer this too, of course.Thanks.
Comments
1) MBT questions
2) Parallel questions
3) Sufficient Assumption questions
For MBT questions, you need to know what must be true, meaning you need to make only valid inferences. For a lot of people, this means drawing out the lawgic to be certain that you're not getting lost in the language.
For parallel questions, you would use lawgic to figure out the structure of the argument so you can cross-reference your answer choices to the lawgical structure of the stimulus.
For sufficient assumption questions, you would use lawgic to figure out any gaps in the argument. A very simple way to do this is:
If the argument says A --B, and concludes that A --> C, then the gap that needs to be filled is B--> C. This makes it really easy to cross-reference your answer choices to fulfill the sufficient gap.
For the remaining question types, you want to spend more time figuring out the conclusion and the support the argument gives to lead you to that conclusion. For some people, it helps them to diagram the argument if the indicators are given. For others, like myself, diagramming takes time away from the argument so I'd rather spend my time working through the argument and defining the argument core (main premise + conclusion). That's when it becomes variable.
This is a great example why you want to hit as many questions as possible while prepping so you become comfortable enough with the questions that you almost instantly know if you should use lawgic or not. Takes some time to get down but it's possible
Beyond that, once every test or so on an LR question there will be an assumption question that has a chain of 3-4 different relationships. I think you need to be safe and lawgic that.