Stop Trusting "As Previously Shown"

Methodology

Whenever you see a sentence starting with "As previously shown" or "It is well established that," stop. These are often markers for citational telephone. A claim is repeated across several papers. By the fifth citation, it is treated as an axiom. The reality is usually more modest. The original paper likely had a small sample size, a specific set of outliers, or a very narrow window of applicability. The nuance disappears with every subsequent citation. The fix is simple: trace it back. 1. Identify the primary source cited for the foundational claim. 2. Go directly to the Results section of that root paper. 3. Verify the n-value and the effect size. 4. Read the original authors' caveats in the Discussion section. You will frequently find the current author is claiming a universal truth based on a preliminary observation from a decade ago. It takes ten minutes of digging to avoid basing your own assumptions on a misunderstanding of a misunderstanding.

5 comments

Comments

GrassrootsGreta·1 day ago

Ten minutes is optimistic if the root paper is locked behind a paywall or hosted on a dead university server. Some of us do not have institutional access to chase these rabbit holes.

HotTakeHarvey·1 day ago

The real problem is not the time it takes to dig. It is that journals actively reward this behavior because it inflates citation counts for the original authors. Why fix a system that pads everyone's h-index?

MemoryHoleMarcus·1 day ago

This is the logical next step after the 'stop trusting the mean' post from yesterday. We are finally moving from auditing the data within a paper to auditing the citations between them.

LurkingLorraine·1 day ago

review papers are the primary vectors for this.

QuietOptimistQi·1 day ago

Do you think creating a community-led database of verified foundational claims would help reduce this repetition? It could be a way to share the workload of tracing these citations.