Abstract Fully-online graduate degree programs are on the rise, generating new questions about how we conceptualize and measure student participation in the virtual classroom. To date, most studies examine participation patterns in single online and/or hybrid courses and do not take into account the demographic characteristics of online students. We develop and test hypotheses that address demographic differences in the nature and intensity of graduate student participation in online-only classrooms for a large degree program in computer science. This work is distinctive because it addresses participation across several classes and across multiple semesters. We select the Piazza forum because it is the required communication mechanism in the program that is the subject of our study. We extract Piazza discussion board activity logs from a sample of 1914 online computer science graduate students, specifically the numerical data indicating the type of access to Piazza students used. We distinguish between active (contributions), passive (viewership) participation and lurking behavior. Given the nature of the dependent variables of interest, we employ different forms of regression analysis. We use logistic regression to address the likelihood of non-participation in the online forum. We then use negative binomial regression to examine the intensity of passive and active engagement, and ordinary least squares regression to examine lurking behavior. We find that the intensity of participation varies by different demographic characteristics, including by age and by race/ethnicity, but not by gender. Our study also shows a notable impact of class size, where increasing class size is associated with decreasing levels of active participation and increasing lurking behavior.

中文翻译：

---

虚拟课堂的潜伏和参与：性别、种族和年龄对计算机科学研究生的影响

摘要 完全在线的研究生学位课程正在兴起，产生了关于我们如何概念化和衡量学生在虚拟课堂中的参与度的新问题。迄今为止，大多数研究检查了单一在线和/或混合课程的参与模式，并且没有考虑在线学生的人口统计特征。我们开发并测试了一些假设，以解决研究生参与计算机科学大型学位课程的纯在线课堂的性质和强度方面的人口统计学差异。这项工作是与众不同的，因为它涉及多个班级和多个学期的参与。我们选择 Piazza 论坛是因为它是我们研究的项目中所需的沟通机制。我们从 1914 名在线计算机科学研究生样本中提取了 Piazza 讨论板活动日志，特别是指示使用 Piazza 学生访问类型的数字数据。我们区分主动（贡献）、被动（观众）参与和潜伏行为。鉴于感兴趣的因变量的性质，我们采用不同形式的回归分析。我们使用逻辑回归来解决不参与在线论坛的可能性。然后，我们使用负二项式回归来检查被动和主动参与的强度，并使用普通最小二乘回归来检查潜伏行为。我们发现参与的强度因不同的人口特征而异，包括年龄和种族/民族，但不因性别而异。