Statistical significance
Use of the word “significant” in scientific writing
One word you see a lot in scientific writing is “significant”. This term has a specific meaning in science. Let’s explore the use of the word “significant”, and how we can ensure it is used correctly in scientific writing.
Statistical significance
The word significant is used in statistics to indicate that a relationship between two variables is not the result of random chance but has a specific cause. In short, statistical significance is a measure of probability. You are calculating the probability that a hypothesis is true, measured against the uncertainty that accompanies a research outcome.
Correct use of the term significance usually means a mathematical analysis has taken place.
Other meanings of significant
However, a mathematical analysis isn’t always essential. For example, in medical writing, the term clinical significance refers to results that have an important clinical effect. This clinical effect could be any improvement to a patient’s function or quality of life, as judged by the patient. The improvement may be determined by both subjective and objective means, and so the results may not agree with tests of statistical significance.
Practical significance describes a research outcome that is meaningful in real life, as judged by the researcher or clinician. The magnitude of the effect is being judged when determining practical significance. So, again, something that has practical significance may not agree with tests of statistical significance.
In colloquial language, the word significant also indicates something of importance, or worthy of attention. It is this use we need to take care with in scientific writing so we don’t confuse what the author is talking about. Is something statistically significant, or is it important and worthy of attention? In this latter case, it may be more appropriate to use another word for significant.
Alternative words
Many journals now discourage the use of significant unless it specifically refers to statistical significance. They suggest using alternative words to indicate when something is important, but not mathematically significant.
A thesaurus can provide a more complete list of alternative words, but some of the most common are:
important
notable
marked
substantial
noteworthy
meaningful
striking
dramatic.
