PrepTest 134

Incorrect. This answer can be tempting because it sounds as if the sample wasn't representative of the general public. However, three issues here. First, we don't know that the injury rate was significantly lower—maybe it was just 0.1% lower, well within the statistical range expected. Second, even if significantly lower, we don't know that that invalidates the comparison between the groups—maybe this group is a bit less injury-prone overall, but the two groups could still be compared. And third, all that aside, this suggests that something in the study affected injury rates—something other than stretching. Since the argument is that stretching doesn't affect injury rates, showing something else does (something about the study) would actually align with that conclusion.

Incorrect. First, there's no indication what proportion of the stretching group is made up by those who had difficulty stretching, so there's not enough force in this answer to weaken the argument. Second, even if some had difficulty stretching, that means that they were still stretching, just with difficulty, so that doesn't undercut the argument.

Incorrect. Without knowing if there was a difference between the two groups in these types of injuries, we can't tell if this affects the argument. That said, (presumably) since the two groups should have the same rate of these injuries, it'd wash out in the comparison and thus have no effect on the argument.

Correct. Argument or Facts:
Argument

Valid or Flawed:
Flawed

Question Type:
Weaken

Stimulus Summary:
Study - Joggers who stretch get as many injuries as those who don't, so stretching doesn't prevent injuries.

Answer Anticipation:
There are two patterns present in this stimulus that both lead us to the same anticipation. And both are related to the central feature of the argument—a study was done. When studies are done, the argument tends to either have a sampling flaw, or a correlation/causation flaw.

So let's start by looking at the sample. Samples need to be large, representative, and random. Here, the stimulus notes that the sample was large, and the premises and conclusion are about people who job, so it's representative. However, it's not random—the scientists found people who were already stretching or not and tracked their injuries. Since the study didn't randomly assign people between the two groups, it's possible that there's something inherently different about the groups. For instance, the group that stretches may be more injury-prone, which is why they stretch to begin with.

And that relates to the second pattern—the anti-causal conclusion. The argument does establish a correlation to show that something doesn't work to prevent a negative outcome. Much like a regular correlation/causation flaw, the correlation isn't enough to prove a lack of causation. This is because the preventative measure might work and actually lowered the negative outcome below what it would have been without it. And this goes back to the sampling flaw—it's possible that the group that stretches would have had a higher injury rate if they hadn't stretched because of some other health conditions.

With both patterns leading to the same answer, let's find one that points out that the group that stretches might be different in that they're more prone to injuries that are prevented through stretching, and thus their injury rate matching the other group actually shows that stretching does work to prevent injuries because that group's injury rate should have been higher.

Answer Explanation:
This answer brings up the intersection between the two flaws we noted. The two groups may have been inherently different, and the stretching group may have been expected to have a higher rate of injury based on how prone they are to injury. If that's the case, then their matching the injury rate of the non-stretching group would actually provide evidence that stretching does work, weakening the argument.

Key Takeaway:
Studys, surveys, and polls should lead you to check for sampling flaws and correlation/causation flaws. For the former, remember that the groups should be randomly determined. For the latter, conclusions that something doesn't have a preventative effect need to compare against what would have happened without the prevention.

Incorrect. The argument is about the number of injuries, not the severity of those injuries, so this answer is out of scope.