Research Tips
Anomalous Results
Sometimes a particular research result lacks face validity: it just does not seem to make sense in light of other results or general observations about whatever is being measured. When this happens, it’s important to try to determine the reason for the anomalous result. Possible explanations include:
- Problems with how a question was asked.
- Problems with the way data were collected.
- Data entry errors.
- Data analysis errors.
- The result is valid and reliable, but just unexpected.
Before accepting the final explanation, it’s important to rule out the others. Even when the other explanations have been ruled out, it’s good practice to look for additional evidence (through replication of the original research or triangulation using other methodologies) to confirm that the anomalous result is correct.
It’s also important to note that this tip applies to results from analysis (based on a group of observations) rather than to a single observation. Single observations that differ greatly from others collected at the same time using the same methodology are often referred to as outliers.