Exercise 2: Sourcing the Bohannon Chocolate Case — Possible Solution
=======================================================================

PRIMARY SOURCE: the original journal article itself, "Chocolate with
High Cocoa Content as a Weight Loss Accelerator," published in the
International Archives of Medicine, reporting the study's own new data
(15 participants, 18 measured variables).

What it could support, even before the hoax was known: a claim strictly
limited to "one small study observed this specific result in this
specific 15-person sample, using this specific design." Even taken at
complete face value, the primary source's own tiny sample size never
supported a general, population-wide claim like "chocolate helps
people lose weight" - that gap existed independent of the later fraud
revelation.

SECONDARY SOURCES: the news articles covering the study - the Daily
Star, Cosmopolitan Germany, and the German and Indian editions of the
Huffington Post, each with headlines like "Slim by Chocolate."

What they could support: at most, "a study exists claiming this
finding" - not "this finding is true" or "this finding is well
established." A responsible secondary source reporting on a single
small study should have noted the sample size and the absence of
independent replication; the real coverage did neither, instead
presenting the finding as settled and actionable ("Why You Must Eat
Chocolate Daily").

The gap in both cases: neither the primary source's own actual data
(a 15-person sample) nor basic secondary-source practice (checking a
study's size and standing before repeating its conclusion) supported
the sweeping claims that actually reached readers.

ANSWER: The primary source was the original journal article, which -
even at face value - could only support a narrow claim about one small,
15-person study, not a general finding. The secondary sources were the
news outlets reporting on it, which could only support "a study made
this claim," not that the claim was true or well-established - yet
they reported it as settled fact.

WHY THIS WORKS AS AN ANSWER
------------------------------
This correctly identifies both source types in the real case, and
explains precisely what claim each one could support versus what claim
actually reached the public - directly matching what the exercise
asked.