The available continuous measurement designs for longitudinal research on perceived usability include immediate and retrospective measurement methods. The present study was conducted to examine the gap between different continuous measurement designs. Thirty users completed a longitudinal study on typesetting software. We used two measurement methods to design three stages of testing. Each stage consists of one immediate measurement and two retrospective measurements. A long-term retrospective test was also conducted and SUS questionnaire measurements were gathered. We observed significant differences in perceived usability between participants with varying quantities of event experiences (i.e. the number of times they used the system). Within short periods of time (3 days), there may be significant differences between retrospective and immediate assessments—especially when users make negative evaluations. In the 3-day period, the user’s perceived usability retrospective evaluation is also based on the most recent experience as opposed to the most unpleasant. The recency effect is lost, however, over a long-term retrospective assessment (1 month later). This study is significant for two main reasons: (i) its unique methodology and (ii) its unprecedented level of consistency between measured results and actual user experiences. We hope that the results presented here provide a workable basis for continuous measurement design in future longitudinal studies. We found that in order to secure accurate measurements, it is necessary to set limits on the measurement time and the number of uses (events).

中文翻译：

---

连续测量方法之间的差距：感知可用性的纵向研究

用于感知可用性纵向研究的可用连续测量设计包括即时和回顾性测量方法。本研究旨在检查不同连续测量设计之间的差距。30 位用户完成了对排版软件的纵向研究。我们使用两种测量方法来设计三个测试阶段。每个阶段包括一个即时测量和两个回顾性测量。还进行了长期回顾性测试，并收集了 SUS 问卷测量结果。我们观察到具有不同数量的事件体验（即他们使用系统的次数）的参与者之间感知可用性的显着差异。在短时间内（3天），回顾性评估和即时评估之间可能存在显着差异，尤其是当用户做出负面评估时。在 3 天的时间里，用户感知可用性的回顾性评估也是基于最近的体验​​，而不是最不愉快的体验。然而，在长期回顾性评估（1 个月后）中，新近效应消失了。这项研究之所以重要，主要有两个原因：(i) 其独特的方法和 (ii) 测量结果与实际用户体验之间前所未有的一致性。我们希望这里提供的结果为未来纵向研究中的连续测量设计提供可行的基础。我们发现，为了确保准确的测量，有必要对测量时间和使用次数（事件）进行限制。