Throughout the term, and particularly near the end of the term, assessing the learning of students becomes a crucial aspect of any learning environment and presents new challenges in the online environment. Changing assessment practices has been difficult for our field, which tends to rely on high-stakes exams to determine course grades. Indeed, when the pandemic first hit, our field became focused mainly on high-stakes testing and how to effectively proctor exams. However, there are other forms of assessment promoted in IMPACT that can be used throughout the term, including diagnostic assessment and formative assessment, that serve to monitor student performance, provide feedback on their learning, and communicate our expectations.

In the shift to the online environment, increased vigilance over academic integrity can come into conflict with the equitable assessment of student learning outcomes. **Do we consider the options for proctoring through an equity lens? Are we actually measuring student learning or privilege? Overall, how can we leverage the online environment to improve assessment practices to become more equitable?**

** **

5 comments

46 views

Recently, I gave a pre-calculus test online and I tried some thing different. Students were given randoms numbers in their problems and if they didn't like the numbers they got then they could get a new randomization. I hope this allowed students the opportunity to get a problem that they could answer. I also hope that it removed my own bias of selecting numbers that are easy for some but hard for others.

Jennifer, Thank you for following up on my summer group exam experiment. I will be taking the same approach next term given the constraints of the pandemic. An example of a "group-worthy" question is at the end so people can see what types of questions I asked. Each problem is worth 10 points graded using rubrics developed by AACU. 5 points are possible for correct and comprehensive concepts used and 5 points for correct and comprehensive calculations. When I compared results on Exam 1, 2, and a comprehensive Final to a past quarter (Fall 2019) when I gave individual exams, the results were comparable. It's important to remember that in the group exam, students can access any resources exam human resources other than me or their group members. In each case, the exam scores were slightly left skewed. But not so much so that the mean and standard deviation are too out of alignment with reality. Here are the results showing the mean and standard deviation:

Group Exam 1 81% (SD = 15)

Individual Exam 1 88% (SD = 10)

Group Exam 2 86% (SD = 14)

Individual Exam 2 81% (SD = 18)

Group Exam 3 - comprehensive final 80% (SD = 12)

Individual Exam 3 - comprehensive final 82% (SD = 16)

Here's an example of a normal curve problem.

Problem 2: Normal Curve

According to a study by the World Health Organization, 1 year old girlsâ€™ heights and weights are normally distributed. The information for each is below

https://www.who.int/childgrowth/standards/Technical_report.pdf

HEIGHT of 1-year-old girls is normally distributedMean height: 59 cm Standard deviation height: 2.1 cm

WEIGHTS of 1-year-old girls are normally distributed:Mean weight: 5.7 kg Standard deviation of weight: 0.13 kg

Use this information to determine the

range of heights and weightsthat would be considered common for a 1-year-old girl. For full credit, clearly state your answer and provide statistical reasons that are comprehensive in using the Empirical Rule and Z scores for normal curvesA 1-year old girl visits the doctor. Her height is 62 centimeters and her weight is 5.4 kilograms. Is this girl common or rare? For full credit, clearly state your answer and provide statistical reasons that are comprehensive in using the Empirical Rule and Z scores for normal curves.