Significance in research results

Significance indicates whether a difference in research results can be explained by chance. A significant difference arises when the difference cannot be explained by chance. If this is the case, valid conclusions can be drawn.

Example of significance

Two measurements of brand awareness were conducted. In January, brand awareness was found to be 20%. For convenience, a margin of error of 3% is used in this example. Normally, the margin of error depends on the sample size and the outcome of the study (the p-value).

In fact, brand awareness is between 17% and 23%. In June, brand awareness appears to be 25%, which means that it could actually be between 22% and 28%. Looking at the measurement alone, there appears to be an increase, but because the margins of error overlap, this cannot be considered a significant difference. In both measurements, brand awareness could also be 23%. A positive development is observed, but due to the margin of error, this cannot be statistically proven.

Significance overlap

If brand awareness were to be 27% in June, this could be considered a significant difference. In this case, awareness is between 24% and 30% and therefore does not overlap with the first measurement. In this case, it would be possible to speak of a significant difference in brand awareness.