Spreadsheet users routinely read, and misread, others' spreadsheets, but literature offers only a high-level understanding of users' comprehension behaviours. This limits our ability to support millions of users in spreadsheet comprehension activities. Therefore, we conducted a think-aloud study of 15 spreadsheet users who read others' spreadsheets as part of their work. With qualitative coding of participants' comprehension needs, strategies and difficulties at 20-second granularity, our study provides the most detailed understanding of spreadsheet comprehension to date. Participants comprehending spreadsheets spent around 40% of their time seeking additional information needed to understand the spreadsheet. These information seeking episodes were tedious: around 50% of participants reported feeling overwhelmed. Moreover, participants often failed to obtain the necessary information and worked instead with guesses about the spreadsheet. Eventually, 12 out of 15 participants decided to go back to the spreadsheet's author for clarifications. Our findings have design implications for reading as well as writing spreadsheets.
https://doi.org/10.1145/3411764.3445634
The ACM CHI Conference on Human Factors in Computing Systems (https://chi2021.acm.org/)