LSAT 90 - Section 4 - Question 11

This is a Most Strongly Supported question.

The stimulus says that in the past, infants who were not breast-fed were fed cow's milk. Then doctors began advising that cow's milk fed to infants should be boiled, as the boiling would sterilize the milk and prevent gastrointestinal infections potentially fatal to infants. And once this advice was widely implemented, there was an alarming increase among infants in the incidence of scurvy, which is caused by vitamin C deficiency. Breast-fed infants, however, did not contract scurvy.

While this is not an ideal experiment, it's what's known as a natural experiment. There are two groups, the "intervention" and "control" groups. The intervention group (boiling cow's milk) exhibited more incidences of scurvy than the control group (breast-fed milk). We suspect that the explanation is that boiling cow's milk was the cause. But how? Why would boiling cow's milk cause the infants to contract scurvy? The stimulus contains another fact: scurvy is caused by vitamin C deficiency. That's a clue about the causal mechanism. Together, the facts (the phenomena) strongly support the hypothesis that boiling cow's milk destroyed vitamin C, which in turn caused scurvy.

Correct Answer Choice (A) says boiled cow's milk makes less vitamin C available to the infant than does the same amount of mother's milk. This is a version of the hypothesis above. That's why it's the right answer. (A) is by no means a “must be true,” but the standard of proof is lower for MSS. You never really reach 100% validity in scientific reasoning anyway. That only happens in formal reasoning.

And notice how similar this is to Resolve Reconcile Explain. (A) would still be correct if this was an RRE question since it would explain the phenomenon in the stimulus. This reveals that the question stem is more or less superficial, and that there is an underlying unity to Logical Reasoning. And here the logic is scientific reasoning. You are presented with a natural experiment that resembles the ideal experiment and asked to come up with a reasonable hypothesis.

What's a natural experiment and why do I put quotations around "control group"? Because natural experiments are less reliable than ideal experiments. That doesn't mean they're not reliable at all. Far from it. Natural experiments can sometimes provide very strong evidence. But they are not ideal because the "control group" didn't control for everything.

Imagine you had an alternative hypothesis, say, the citrus shortage hypothesis. Under that hypothesis, there was a shortage in citrus fruits that just happened to coincide with the intervention (boiling cow's milk). Perhaps it's actually the citrus shortage that caused the increase in scurvy.

Okay, but we can preclude this hypothesis with an ideal experiment. We can run a control group. Hold everything else equal (including access to citrus) and feed the control group breast milk. If they don't develop scurvy, then it can't be the citrus shortage that caused the scurvy. That's kind of like what happened in the natural experiment of the stimulus. The difference is that we have no assurances that the breast-fed babies' exposure to citrus was the same as the intervention babies' exposure. It's possible that breast-fed infants somehow had priority access to oranges. It's just very unlikely. That distance is there, though, in the natural experiment, whereas in the ideal experiment, we have assurances that the control group did in fact control for everything. That's why natural experiments are weaker than ideal ones. How much weaker? That depends in turn on which hypotheses you're trying to eliminate. For the citrus shortage hypothesis, it's highly unlikely that the breast-fed group would have had priority access and, therefore, the natural experiment is not that much weaker than the ideal experiment.

Answer Choice (B) says infants who consume cow's milk that has not been boiled frequently contract potentially fatal gastrointestinal infections. (B) could have been correct if it had said sometimes, potentially, or even in danger of contracting. Doctors recommended the intervention precisely to prevent gastrointestinal infections, so it is a reasonable assumption that they thought of unboiled cow’s milk as an actual risk. But “frequently” is too strong to be supported.

Answer Choice (C) says mother's milk can cause gastrointestinal infections in infants. This is just a mishmash. We do know that unboiled cow's milk potentially can cause gastrointestinal infections, but can mother's milk cause it? Nothing in the stimulus suggests the answer is one way or another.

Answer Choice (D) is really attractive. If you picked (D), you probably read some alternate version of (D) that resembles the following: when doctors advised that cow's milk fed to infants be boiled, they did not know that this intervention would lead to vitamin C deficiency, which then would lead to scurvy.

Then (D) would be pretty well supported. It seems clear that when doctors began advising to boil the milk, they did not anticipate that scurvy would be a consequence. I am guessing they also did not know that infants depended on milk for their vitamin C, because if they had some other source, for example, like orange juice, who cares about the vitamin C in the milk? Just drink your orange juice.

But that is not what (D) says. (D) just says that when doctors advised cow's milk to be boiled, the cause of scurvy was a mystery. This I am not so sure about. It is totally possible that doctors knew the cause of scurvy. Lots of sailors knew vitamin C deficiency was the cause of scurvy for hundreds of years. They just did not think this intervention (boiling cow's milk) would result in vitamin C deficiency and hence scurvy.

Answer Choice (E) says that when doctors advised that cow's milk fed to infants be boiled, most mothers did not breast-feed their infants. We know that when this intervention happened, there were some infants who were drinking cow's milk and some who were breast-fed. But nothing in the stimulus says which set is larger, so (E) is totally unsupported.