Checking sources against each other
Corroborate: check whether multiple sources agree on the same facts and investigate why they might not.
What a learner can do afterwards
- Wineburg corroboration heuristic
- Historical Thinking in the Elementary Years (ERIC)
1 · Read
A claim is something someone says is true. Corroboration checks a claim against other clues about the same thing. Put accounts side by side. List each claim, mark matches and differences, then check the clues behind them. Counting matching stories is not enough.
Independent accounts do not get their story just by copying one another. Matching details from people who could see an event and made separate records can add support, but they can still share a mistake. Check a less common account too: its writer may have seen something others missed.
An internal check looks within one source for statements that clash. An external check compares other evidence. To check a bridge date in a book, seek a separate bridge record and ask whether it copied the book. A diary may leave parts out. Read carefully, then compare relevant records to test details and find omissions. Age and source type do not make an account complete.
To check an event's order, seek a record listing it. A match record of goals by half addresses when a goal happened. School plans and fair programs show intended order and times. Compare records of what happened too. A tied count needs clues, not a deciding vote. Check where each record's details came from.
Bar charts count reports in groups. Read each bar's name and height. To find how many more, subtract: 8 - 4 = 4 more reports. To find the largest group, choose the tallest bar: among 8, 4 and 2, it has 8 reports. Report counts do not decide what happened.
To find a chart's total, add its groups. Count on from 4: add 3 to reach 7, then 2 to reach 9. So 4 + 3 + 2 = 9 sources. Count each source once, not in two groups.
An estimate is approximate, not an exact measurement. Different estimates need not mean lying. Confidence or averaging guesses does not make a measurement. Seek measured clues. Matching separate records from observers may support a flood's start day while its depth estimates differ. Check the depth further. Disagreement does not make every clue useless, and agreement on one detail does not settle another.
Coverage means the people, places and times described. Check whether lunch accounts mean the same day and group. Dates may refer to events, writing or sending. Four accounts copying an invitation's time repeat one plan; a clock record of arrivals reports what happened. Check what each time means and where it came from before deciding the accounts clash.
To check when a building was finished, look for a record of that building's completion. Calling it new does not give an exact completion date. Check that the record concerns the same building.
A message may give a time without showing how it was known. A checked video of a meeting with a visible clock may add a recorded time. Check that it shows the same event and that the clock is right.
Before deciding, investigate differences and trace where details came from. Matching words may come from copying. Differences may concern different scenes or mistakes. Check what each writer could know. If accounts conflict about the same detail and no clue settles it, report that detail as uncertain.
Compare accounts about the same thing. Check where their details came from. More matching words do not always mean more evidence.
This reading was independently written for this dataset.
2 · Watch
Take it off screen
Where it sits
Where this leads
Jobs that lean on this skill. Follow one to see everything it is built on.
24 questions wait behind this lesson, each with its answer explained. Every answer feeds the sky: stars light as they are learned, and dim when it is time to come back.