🤷🏿 🤷🏼‍♂️

If you’re looking for differences and can’t find any, this can mean one of two things:

  1. There is a large amount of overlap between language sets, and thus no statistically significant differences between them

  2. One or both of the language sets are too small to retrieve differences that are statistically significant 

What do Relative Insight consider to be statistically significant differences?

To ensure that the differences surfaced by the technology are not just happening by chance, the platform will only display differences that have a 99%+ confidence rating.

Small language sets make it difficult for the platform to make judgements that will be statistically valid.  While Relative Insight can analyse language sets as small as 200 words, we recommend uploading data sets of at least 2,000 words to get meaningful insights. 

Did this answer your question?