Replication
Replication is repeating an investigation, ideally by other scientists, to see whether the same procedure gives the same result.
Definition
Replication means doing the whole experiment again to see whether you get the same answer; if someone else can repeat it and get the same result, the result is trustworthy. It differs from running multiple trials, which are repeats within one experiment: replication repeats the entire investigation, often by different people with different equipment, either under the same conditions (direct replication) or under deliberately varied ones (conceptual replication). Successful replication rules out chance, undisclosed procedural quirks, and equipment-specific artifacts, and it is the strongest evidence that a finding is real. A failure to replicate does not automatically mean the original was wrong; it triggers a search for the uncontrolled difference.
Example
A group finds that salt water freezes at a lower temperature than plain water, and another group follows the steps exactly and gets the same result. Group A reports a mean dissolving time of $26\ \text{s}$ at $30^\circ\text{C}$, and Group B, with its own stopwatch, salt, and beaker, gets $28\ \text{s}$, agreeing within the range each group saw. Group A measures a rock's density as $2.7\ \text{g/cm}^3$ while Group B gets $2.5\ \text{g/cm}^3$; investigation shows Group B read the meniscus from above and overstated the volume, so the replication failure exposed a technique error rather than a property of the rock.
Key Insight
A result that happens only once might be luck or a mistake; a result that happens every time, for everyone, is science. Replication is only possible when the procedure is written clearly enough for someone else to follow, so vague methods make results impossible to check. Science does not run on trust, it runs on replication, and a finding that only its discoverer can produce is at best not yet a finding.