A small check for
a changed file.
A familiar name does not guarantee familiar contents. A checksum gives you another way to compare.
An educational explanation of a general technology concept.
Read the contents, not the label
A checksum is a value calculated from data using a defined procedure. For file verification, a cryptographic hash such as SHA-256 produces a fixed-length result from the file's contents. Renaming a file does not, by itself, change those contents.
The same algorithm applied to the same bytes produces the same result. If the results differ, the inputs differ. This makes a hash useful when a file moves between devices or is downloaded from a source that publishes an expected value.
Compare like with like
In a hypothetical example, a community group shares a printable meeting agenda. A reader downloads it, calculates its SHA-256 value and compares that value with one supplied through a trusted channel. The reader must compare the same file version using the same algorithm.
Calculate valueReceived file
Calculate value
If the values differ, the reader pauses and obtains the file again. The difference alone does not explain whether the cause was an incomplete transfer, an intentional revision or another change. A checksum detects a difference; it does not diagnose it.
A match has limits
A matching hash is only as helpful as the expected value and the algorithm. If someone replaces both a file and its published hash, comparing the two cannot establish who created the file. Authenticity requires a trustworthy reference or another verification mechanism.
Different inputs can theoretically share a hash, a situation called a collision. Modern cryptographic algorithms make deliberate collisions difficult, but a match is not a promise of absolute certainty. Nor does a hash show that a document is accurate, safe to open or suitable for a particular purpose.
Keep the questions separate: did the data change, where did it come from, and should it be trusted?