Weighing Historical Evidence: Bias, Corroboration and Gaps

No single document tells the whole truth about the past. Every source was made by someone, for some purpose, and survived for reasons that may have nothing to do with how accurate it is. The work of history is not finding the one perfect source but weighing many imperfect ones.

The American Historical Association (AHA) states it plainly in its Statement on Standards of Professional Conduct: "Multiple, conflicting perspectives are among the truths of history." That does not mean every claim is equally good. It means that reaching a sound conclusion takes method.

This guide sets out that method in four parts: understanding a source, reading its bias, checking it against others, and accounting for what is missing.

Start by meeting the source

The U.S. National Archives teaches document analysis in four steps that apply to any kind of source, from a letter to a photograph to a sound recording:

1. Meet the document.
2. Observe its parts.
3. Try to make sense of it.
4. Use it as historical evidence.

The order matters. Many mistakes come from jumping straight to step four, quoting a striking line before asking what kind of document it came from. The Digital Inquiry Group, which publishes the Reading Like a Historian curriculum, frames similar skills as reading strategies, among them sourcing, contextualizing, corroborating and close reading.

Sourcing: who, when, why and for whom

Sourcing means asking where a document came from before deciding what it proves. The Library of Congress suggests questions such as: What was the creator's purpose in making it? Who was the audience? What was happening at the time? What biases or stereotypes appear in it?

Closeness to events helps, but it is not a guarantee. The Society of American Archivists notes in its Dictionary of Archives Terminology that the idea of a primary source rests on "the belief that firsthand accounts are more accurate," and it quotes a reminder that primary sources are subjective interpretations that historians should approach "with a healthy skepticism." An eyewitness may have seen only part of an event, misunderstood it, or had reasons to shade the story.

Bias is information, not just a defect

Every source has a point of view. The National Archives offers teachers a separate worksheet for understanding perspective, designed to show how a creator's background, beliefs and experiences shape their point of view.

The useful question is not "is this source biased?" but "biased in what way, and what does that tell me?" Imagine two accounts of a hypothetical 1910 factory strike: a company's annual report and a pamphlet printed by the strikers. Each will stress different facts and leave others out. Read together, they show what each side wanted the public to believe, and where they agree, despite opposing interests, you may be on firmer ground. A source can be unreliable about one thing and excellent evidence for another, such as its author's own attitudes.

Corroboration: checking sources against each other

Corroboration means comparing a source with others. The Library of Congress's guidance for teachers suggests asking whether a source agrees with other primary sources and looking for other sources that offer "support or contradiction."

A few habits make corroboration meaningful:

• Check independence. Two newspapers telling the same story may both be reprinting one original report. Agreement between sources that copied each other is not two pieces of evidence.
• Compare different kinds of evidence. A diary, a tax record, a photograph and a radiocarbon date fail in different ways, so agreement among them is more persuasive. See our guide to radiocarbon dating for how scientific dates carry their own uncertainty.
• Take contradictions seriously. When sources disagree, the disagreement itself is evidence about who knew what and who wanted to say what.

Gaps and silences

What is missing can matter as much as what survives. The Guidelines for Primary Source Literacy, approved in 2018 by the Society of American Archivists and the Association of College and Research Libraries, urge researchers to think critically about "what sources were never created, what sources may no longer exist, and what sources are collected." Some people were never asked to write things down; some records were thrown away; some were never thought worth keeping.

Scale matters too. The National Archives says that only 1 to 3 percent of the documents and materials created in the course of federal government business are kept forever.

Even the story of a gap needs checking. A 1996 article in the National Archives' Prologue magazine by Kellee Blake examined the 1890 U.S. population census, which reference sources, it noted, "routinely dismiss" as "destroyed by fire" in 1921. The real story was more complicated. After the fire on January 10, 1921, the Census director estimated that 25 percent of the 1890 schedules had been destroyed and half of the rest damaged by water, smoke and fire. Many records survived the blaze itself. In December 1932 the Census Bureau listed the original 1890 population schedules among papers scheduled for disposal, and Congress authorized their destruction on February 21, 1933. The article notes reports that they were destroyed in 1934 or 1935. Fragments survive, with no fewer than 6,160 names indexed, and the article advises researchers to verify that the records they seek did not survive before giving up.

The lesson reaches beyond genealogy. An absence in the record has its own history, and arguments built on silence ("there is no record of it, so it did not happen") need to show that a record would have been created and kept in the first place.

Holding conclusions with the right confidence

Weighing evidence ends in a judgment, not a certainty. The AHA describes professional historians as "as interested in defining the limits and uncertainties of their own arguments as they are in persuading others that those arguments are correct."

In your own work, that means:

• matching your language to your evidence, using words like "probably" or "the surviving records suggest" when that is all the evidence supports
• saying where historians disagree, and why
• citing your sources fully, so others can follow what the AHA calls the "trail of evidence" and check your reasoning

The AHA is equally clear about the floor beneath all of this: historians "do not fabricate evidence," and those who alter or ignore evidence lose the trust that makes their work useful.

Key takeaways

• Analyze a source before using it: who made it, when, why and for whom.
• Bias is not a reason to discard a source; it is information about the creator.
• Corroborate with independent sources, ideally of different kinds.
• Account for gaps: what was never recorded, what was lost, and what was never kept.
• State conclusions with the confidence the evidence supports, and show your sources.

Sources

• American Historical Association, Statement on Standards of Professional Conduct
• U.S. National Archives, Document Analysis Worksheets
• Digital Inquiry Group, Reading Like a Historian: History Lessons
• Library of Congress, Getting Started with Primary Sources
• Society of American Archivists, Dictionary of Archives Terminology: primary source
• SAA and ACRL/RBMS Joint Task Force, Guidelines for Primary Source Literacy (June 2018)
• Kellee Blake, "First in the Path of the Firemen": The Fate of the 1890 Population Census, Prologue (Spring 1996)
• U.S. National Archives, Getting Started Overview