Bike-Lane Pilot: Which Inference Is Supported?
City planner Omar is reading a pilot evaluation. The report states: 'After a 6-month protected bike-lane pilot on Elm Street, reported cyclist injuries fell from 18 to 7, while cyclist counts rose 22%. The city also installed brighter streetlights in month 3. Nearby Oak Street, without a lane, saw injuries rise from 9 to 11. The report concludes that the lane made cycling safer.' Omar must answer inference items. He notices the report uses reported injuries, not hospital records, and that the streetlight installation overlapped with the pilot. A colleague claims the data prove the lane reduced injuries by 61%. Omar has to decide which conclusions are supported, which are merely consistent, and which go beyond the passage.
1. Which conclusion is best supported: (A) the lane reduced injuries by 61%; (B) the lane is associated with a larger drop in reported injuries on Elm than on Oak, but the cause is not isolated; or (C) the streetlights caused the drop? Explain.
Read the worked answer
B is best supported. The passage gives a before-and-after drop on Elm and a comparison with Oak, which did not get the lane. That pattern is consistent with a safety benefit. But the report also installed brighter streetlights during the pilot, and it measures reported injuries rather than hospital records. Those details prevent a clean causal conclusion.
A is arithmetically tempting: 18 to 7 is a 61% drop. But the percentage is based on reported injuries, and the design does not isolate the lane from the streetlights. C is also unsupported: the passage does not separate the streetlights' effect from the lane's effect. B uses the cautious language the data allow: association plus an unresolved confound.
2. A colleague says 'the data prove the lane made cycling safer.' What is the strongest objection available from the passage?
Read the worked answer
The strongest objection is that the evaluation is observational and confounded. The streetlights were installed in month 3, during the same pilot, so any change on Elm could reflect lighting rather than the lane. In addition, the outcome is 'reported cyclist injuries,' not a consistent hospital or police record, so reporting behavior may have changed.
There is also an exposure issue: cyclist counts rose 22%. A rise in riders could change both the number of injuries and the rate per rider. The passage supports a conclusion that the lane is associated with a favorable pattern, especially compared with Oak, but 'prove' is too strong. A good answer names the confound and the measurement limit rather than simply rejecting the report.
3. Which inference is unwarranted: 'the lane may have encouraged more cycling' or 'the lane will reduce injuries citywide'? Explain.
Read the worked answer
'The lane will reduce injuries citywide' is unwarranted. The passage covers one street over six months, with a confound and reported-injury data. Extending the result to the whole city assumes that Elm's traffic, riders, and reporting conditions generalize, which the passage does not establish.
'May have encouraged more cycling' is weaker and more consistent with the 22% rise in cyclist counts, though the passage does not prove that the lane caused the increase. The word 'may' keeps it as a possible inference. The citywide claim is a prediction stated with certainty from a single-site pilot, so it goes beyond the evidence.