The methodology

If you’re interested, here’s more about what we did…

The methodology

We explored how users read Turtl Docs compared to webpages based on the findings of a research study by Nielsen Norman Group and our own data.

Our initial dataset consisted of 24,109 documents. And, similar to the Nielsen research, sessions that lasted less than four seconds and more than 10 minutes, were removed.

Our own data was then further normalized to reduce the impact of outliers.

To do this, we removed documents that had less than 20 words on a page, and removed the top and bottom 10% of documents based on the number of readers, the average read times, and the number of pages – bringing our final normalized dataset to 10,831 Turtl Docs and 1,340,448 user sessions.

Analysis of our data used a best-fit formula to plot the average reading time for Turtl Docs of different word counts, comparing the results from the Nielsen Norman study on webpages.