PhD Candidate, University College London Interaction Centre
How do finance workers complete data entry work to make sure you are paid in time?
Think back to the last time you came back from a conference. You probably saved all your receipts from the trip, and had to enter them into a computer-based accounting system to claim back your expenses. You probably started out by entering the items and prices into the system. Maybe you needed to open a website to convert prices from one currency to another. After entering the receipts, you might then have needed to enter your bank details. This might require a further forage for this information from an online banking system. After this, entering the correct budget code might require further searching through old emails to find the correct information. Throughout this scenario, how many times have you interrupted the data entry task to go and look up relevant information? Alternatively, you might have gathered all the required information beforehand, to not have to interrupt the data entry task. Maybe, you entered information you knew first, and left the items you had to spend time looking up until the end? Whatever way you chose to complete this task, it involved making decisions as to when to gather relevant information to enter it into the computing system.
People often have to go in and out of applications, documents and even physical locations during routine data entry work, which can be disruptive and increases likelihood of errors. It is challenging for users to organise and manage their information in a structured way.
My research is concerned with how people in finance offices collect and organise information as part of their data entry work. I have conducted two field studies interviewing employees and observing them doing their work. This showed that people carefully plan and organise their paper sources, such as receipts. Digital sources, such as Excel spreadsheets, were fetched in an unplanned and reactive manner as and when they were needed. I then conducted three experiments to see how time costs to collect information influences people's strategies. This showed that people group items that have a similar costs together, rather than finishing one data entry task. The consequence is that people collect and enter all easy-to-find items first, even if that means interleaving between tasks.
For the final part of my PhD, I plan to design a prototype to better support people in their information needs, and evaluate these with end users.
Abstract: Nurses frequently have to program infusion pumps to deliver a prescribed quantity of drug over time. Occasional errors are made in the performance of this routine number entry task, resulting in patients receiving the incorrect dose of a drug. While many of these number entry errors are inconsequential, others are not; infusing 100 ml of a drug instead of 10 ml can be fatal. This paper investigates whether a supplementary graphical number representation, depicting the magnitude of a number, can help people detect number entry errors. An experiment was conducted in which 48 participants had to enter numbers from a ‘prescription sheet’ to a computer interface using a keyboard. The graphical representation was supplementary and was shown both on the ‘prescription sheet’ and the device interface. Results show that while overall more errors were made when the graphical representation was visible, the graphical representation helped participants to detect larger number entry errors (i.e., those that were out by at least an order of magnitude). This work suggests that a graphical number entry system that visualizes magnitude of number can help people detect serious number entry errors.
Pub.: 27 Oct '15, Pinned: 25 Aug '17
Abstract: During data entry tasks, small errors can result in catastrophe, for instance adding an extra zero to a drug dose when programming an infusion in a hospital. For this reason understanding users’ error checking behavior is highly important. One aspect that can affect error checking is the interface that a user must interact with to enter data. Often user interaction with interfaces is evaluated based on speed or error rate. In this paper, in addition to this, we also explore how different types of interface can affect a user’s error checking behavior in a multitasking environment. We show that a fast to use and familiar interface discourages users from carrying out thorough visual checking in a number transcription task. We also found that having participants perform an additional secondary task while doing the number entry task made participants less likely to check the inputted numbers for errors.
Pub.: 27 Oct '15, Pinned: 25 Aug '17
Abstract: Interruptions are disruptive because they take time to recover from, in the form of a resumption lag, and lead to an increase in the likelihood of errors being made. Despite an abundance of work investigating the effect of interruptions on routine task performance, little is known about whether there is a link between how quickly a task is resumed following an interruption (i.e., the duration of the postinterruption resumption lag) and the likelihood that an error is made. Two experiments are reported in which participants were interrupted by a cognitively demanding secondary mental arithmetic task while working on a routine sequential data-entry task. In Experiment 1 the time-cost of making an error on the primary task was varied between conditions. When errors were associated with a high time-cost penalty, participants made fewer errors and resumed the primary task more slowly than when errors were associated with a low time-cost penalty. In Experiment 2 participants were prohibited from resuming the primary task quickly by a 10-s system lockout period following the completion of the interrupting task. This lockout period led to a significant reduction in resumption errors because the lockout prohibited fast, inaccurate task resumptions. Taken together, our results suggest that longer resumption lags following an interruption are beneficial in terms of reducing the likelihood of errors being made. We discuss the practical implications of how systems might be designed to encourage more reflective task resumption behavior in situations where interruptions are commonplace. (PsycINFO Database Record (c) 2013 APA, all rights reserved).
Pub.: 26 Jun '13, Pinned: 25 Aug '17
Abstract: Link-shortening services save space and make the manual entry of URLs less onerous. Short links are often included on printed materials so that people using mobile devices can quickly enter URLs. Although mobile transcription is a common use-case, link-shortening services generate output that is poorly suited to entry on mobile devices: links often contain numbers and capital letters that require time consuming mode switches on touch screen keyboards. With the aid of computational modeling, we identified problems with the output of a link-shortening service, bit.ly. Based on the results of this modeling, we hypothesized that longer links that are optimized for input on mobile keyboards would improve link entry speeds compared to shorter links that required keyboard mode switches. We conducted a human performance study that confirmed this hypothesis. Finally, we applied our method to a selection of different non-word mobile data-entry tasks. This work illustrates the need for service design to fit the constraints of the devices people use to consume services.
Pub.: 28 Jul '16, Pinned: 25 Aug '17