Article Preview
TopIntroduction
Spreadsheets play an increasingly important role in the operations of companies and other organizations. Projections indicate that by 2012, over 50 million people in American workplaces will create spreadsheets or databases as part of their everyday work (Scaffidi, Shaw, & Myers, 2005). Worldwide, over 500 million people use spreadsheets (Microsoft, 2010). This popularity stems from spreadsheets’ usefulness for a broad range of tasks that include organizing, analyzing, exploring, and visualizing data (Nardi, 1993). In prior studies, we have also found that spreadsheets are becoming a crucial resource for storing and manipulating data that information workers retrieve from web sites (Scaffidi, Cypher, Elbaum, Koesnandar, & Myers, 2007).
With the rising importance of spreadsheets in the organizational context, it is vital to examine the problems reported when people use editors such as Microsoft Excel to create and modify spreadsheets. One reason is that the acceptance and adoption of technology is strongly controlled by its perceived usefulness, as well as its perceived ease of use (Davis, 1989; Ma & Liu, 2004). Therefore, understanding the relative strengths and weaknesses of different spreadsheet editors is likely to be useful for guiding organizations as they select a tool for their use. In addition, identifying common questions can provide researchers with insights about where spreadsheet editors as a whole need improvements, in order to facilitate smoother and more productive adoption.
Moreover, it is important to characterize the problems that people ask for help with because these problems directly affect the support that organizations will need to provide to users. In particular, an assessment of training needs—including an identification of problems commonly reported by users—is a necessary precursor to implementation and testing of training materials (Bostrom, Olfman, & Sein, 1998; Oinas-Kukkonen, Hohtari, & Pekkola, 2010). Organizations typically respond to requests for help through an information technology department or through informal peer support by users who are often called “local developers” or “gardeners” (Nardi, 1993; Morch & Asand, 2006). These IT departments and gardeners invest time in learning how to overcome problems that people commonly ask for help with solving, in order that they can provide support to problems as needed without having to discover solutions during a crisis. Making an informed decision about this investment of time requires knowing what kinds of problems users commonly ask for help with solving.
To date, empirical studies have focused on only a narrow range of problems reported by spreadsheet users. In particular, despite the fact that spreadsheet editors have progressed far beyond simple calculation engines, formula errors are the only spreadsheet user problem that has received extensive investigation. Studies have shown that formula errors are common (Panko, 1998; Powell, Baker, & Lawson, 2009) and can lead to monetary loss (Caulkins, Morrison, & Weidemann, 2007). In response, researchers have proposed a variety of approaches for improving the process by which spreadsheets are created and tested (e.g., Panko, 1999; Rothermel, Cook, Burnett, Schonfeld, Green, & Rothermel, 2002; Panko, 2007). Yet formula errors are only one of many conceivable problems that users might be facing, and it is not at all clear that they remain the most important category of problem from the standpoint of designing training materials and support systems.
In response to this gap in our understanding of common requests for help, we conducted a preliminary study in prior work, where we investigated questions that users posted to an online forum about Microsoft Excel (Chambers & Scaffidi, 2010). We identified seven common categories of problems and noted that only 18% of posted user questions were about formulas.