Observations on Typing from 136 Million Keystrokes

What can we learn from analyzing 136 million keystrokes?

Overlapping keypresses (rollover) are suprisingly common and it can indicate faster typing.

There is a lot of variation in individual typing styles but most people can be characterized by one of eight typist categories.
Faster typists make generally less errors. Slower typist especially replace letters with wrong ones when making typing errors.

Abstract

We report on typing behaviour and performance of 168,000 volunteers in an online study. The large dataset allows detailed statistical analyses of keystroking patterns, linking them to typing performance. Besides reporting distributions and confirming some earlier findings, we report two new findings. First, letter pairs typed by different hands or fingers are more predictive of typing speed than, for example, letter repetitions. Second, rollover-typing, wherein the next key is pressed before the previous one is released, is surprisingly prevalent. Notwithstanding considerable variation in typing patterns, unsupervised clustering using normalised inter-key intervals reveals that most users can be divided into eight groups of typists that differ in performance, accuracy, hand and finger usage, and rollover. The code and dataset are released for scientific use.

Faster typists make use of rollover where keypresses are overlapping. Click here to download the full video.

The 136M Keystrokes Dataset

The data was collected using an online typing test following scientific standards for testing typing performance. The test was published on a free typing speed assessment website. Users transcribed sentences voluntarily after giving their informed consent that their anonymized data will be collected and used for research purposes.

Data

Text files of the keystrokes in a .zip file (1.4 GB zipped, 16 GB unzipped). The data contains the keystroke-by-keystroke entries typed by every participant. The package also includes a metadata file. See readme.txt for introduction and license. This data is free to use for research and non-commercial use with attribution to the authors.

Code

Code for capturing the keystroke data in the online typing test.

Publication

PDF, 2.8 MB
Dhakal, V., Feit, A., Kristensson, P., O., & Oulasvirta, A. 2017. Observations on Typing from 136 Million Keystrokes. In Proceedings of the 2018 CHI Conference on Human Factors in Computing Systems (CHI ’18).

Best paper honorable mention

	    
@inproceedings{dhakal2018observations,
author = {Dhakal, Vivek and Feit, Anna and Kristensson, Per Ola and Oulasvirta, Antti},
booktitle = {Proceedings of the 2018 CHI Conference on Human Factors in Computing Systems (CHI '18)},
title = {{Observations on Typing from 136 Million Keystrokes}},
year = {2018}
publisher = {ACM}
doi = {https://doi.org/10.1145/3173574.3174220}
keywords = {text entry, modern typing behavior, large-scale study}
}

Contact

For questions and further information, please contact:

Antti Oulasvirta

Email:
antti.oulasvirta (at) aalto.fi

Acknowledgements: This work was funded by the European Research Council (ERC; grant agreement 637991) and EPSRC (EP/N010558/1 and EP/N014278/1). Data collection was supported by Typing Master, Inc. We thank Samuli De Pascale for programming support.