Self-study
Support Records reveal that of physical therapy patients who received less than six weeks of treatment, about 31 percent showed major improvement, regardless of whether they were treated by a general practitioner or by a specialist. ██ ████████ ███ ████████ ████████ ███████ ███ █ ██████ █████ █████ ██████████ ██ ███████ ████ ████ ███████ ██ █ ███████ ████████████ ██ ██ █ ███████████ █████ ██ ███████ ██████ █████ ████████████ ██████████ ███ ██████ ███████ ██████ █ ██████████ ██ █ ███████ ████████████ ███ █████████ ████████ ███████ ████ ███ ██████ █████ ███████ ██ █████ ████████████
The Phenomenon
Two facts about physical therapy patients, broken out by treatment length:
Short-term patients (less than six weeks): about 31% showed major improvement. This rate was the same for patients treated by a GP and patients treated by a specialist.
Long-term patients (six weeks or more): about 50% showed major improvement. Again, the rate was identical for GP patients and specialist patients.
It's worth pausing on the phrase
These matching rates are the phenomenon. The author observes them and offers a hypothesis to explain them: since the rates are the same no matter which practitioner patients saw, the choice between a GP and a specialist must not affect your chances of major improvement.
Equal Rates ≠ Equal Outcomes
The author treats the matching improvement rates as proof that the choice doesn't matter. But matching rates can be misleading if the groups being compared aren't similar.
Here's an analogy. Suppose 60% of brain surgery patients and 60% of podiatry patients show major improvement. Same rate. But you wouldn't conclude that it doesn't matter whether you see a brain surgeon or a foot doctor, because the patients going to each doctor have completely different problems. Send a brain injury patient to a foot doctor, and their recovery rate probably drops to near zero.
The same issue applies here. The patients seeing GPs might be a relevantly different from the patients seeing specialists. If they are, then matching overall rates don't tell us anything about whether the choice matters for a given patient.
What the author sees:
GP
31%
Specialist
31%
↑ "Same rate, so the choice must not matter."
What could be hiding underneath:
GP
Group X
50%
Group Y
10%
Specialist
Group X
10%
Group Y
50%
(Hypothetical numbers to illustrate the flawed thinking.)
In this scenario, both practitioners show 31% overall, but the GP is far more effective for Group X patients while the specialist is far more effective for Group Y patients. A Group Y patient who goes to a GP would have only a 10% chance of major improvement. The choice of practitioner would matter enormously for that patient, even though the aggregate numbers are identical.
The two time periods (short-term and long-term) are a bit of a distraction. They're included to reinforce that the rates match across both groups. But this doesn't fix the core problem. Even if the rates match across every time period, the choice could still matter if the patient populations seeing each practitioner differ in some systematic way.
We're looking for an answer that identifies something the argument fails to account for about the comparison between GP patients and specialist patients.
18.
The reasoning in the argument ██ ████ ██████████ ██ █████████ ██ ███ ███████ ████ ███ ████████
a
presumes, without providing ██████████████ ████ █████████████ ██ █████████ █████████████ ██ ████████ █████ █████ ███████████ ██████ ██████ ██ ███ ██ █████ █████████████ ██ ████████ █████ ███ ███████████ ████ ███ ██████
Both the premises and the conclusion are about major improvement. The 31% and 50% figures specifically measure major improvement, and the conclusion is about major improvement. So the argument never makes a leap from "any improvement" to "major improvement." (A) would describe a flaw if the premises tracked improvement generally and then the conclusion drew a claim about major improvement specifically. That's not the structure of this argument.
b
provides no information █████ ███ █████ ██ ████████ ████ ███████ ██████████ ██ ███████ ██ █████████ █████████
The argument doesn't need information about which injuries require short-term vs. long-term treatment. The two time periods are presented to show that GP and specialist recovery rates match across both groups. That's all they're doing. The relevant question isn't what kinds of injuries distinguish short-term patients from long-term patients. It's whether the GP patients within each time period are comparable to the specialist patients within that same time period. More detail about what injuries fall into which time category wouldn't address that question.
c
overlooks the possibility ████ ████████ ███ ████ ████████ ██████ ██ ██████ █████████ ██ ███ ██ ███ ███ █████ ██ ███████ █████████████ ████ ██ ███ █████
The statistics come from
d
fails to indicate ███████ ███ ██████ ██ ████████ ████████ ███ ███ █ ███████ ████████████ ███ █████ ██ ███ ██████ ███ ███ █ ██████████
We already know the percentage of patients who showed major improvement for each practitioner type. Once you have the percentages, the raw number of patients in each group doesn't change the comparison. If 31% of GP patients improved and 31% of specialist patients improved, those rates are equal whether 500 people saw a GP or 500,000 did.
e
overlooks the possibility ████ ███████████ ███ ███████ █████████████ ████ ████ ██ █████ ██ ████████ █ █████████ ████ ██ ██████
If each practitioner excels at a different type of injury, then the choice of practitioner could matter for an individual patient, even though the aggregate rates are identical. A patient whose injury falls into the specialist's area of strength would have a higher chance of major improvement by choosing a specialist. The reverse would be true for a patient whose injury is better suited to a GP. The matching overall rates could simply reflect each practitioner excelling with their own typical patient base, not that the choice is irrelevant.