Assessment Report

Course: STAT 217 Intermediate Statistics
Semester: Spring 2019
Submitted by: Mark Greenwood and Greta Linse

 

Overview

Number of students in course: 185
Number of students assessed: 18

Learning Outcomes

LO# Learning Outcomes
1. Interpret and draw inferences from mathematical or statistical models represented as formulas, graphs, or tables.
2. Represent mathematical or statistical information numerically and visually.
3. Employ quantitative methods such as arithmetic, algebra, geometry, or statistical inference to solve problems.

Assessment Questions

Aligned LO Projects

 

1-3

 A project where the students were tasked to fit multiple models and check the model assumptions and compare the different models. This was the third project of this format in the class and occurred near the end of the semester and they were required to work in groups to complete the assignment. The students ran the statistical software R to get their results and then wrote a report containing their results. They used a real data set on weights of bears to consider building a model to predict bear weight (which is hard to measure) from a suite of other variables that are easier to measure.

Assessment

Criteria for Learning Outcomes LO 1 LO 2 LO 3
Total number assessed 18 18 18
Number of students demonstrated acceptable level 16 18 12
Proportion of students rated as acceptable 89% 100% 67%
Does this meet minimum 2/3 threshold? yes yes at threshold
Comments/ideas for better alignment of course or assignment Most students were able to correctly interpret the graphs that they made in the project. The students were very capable at getting the required results from the statistical software and including them appropriately in the report. Students struggled to correctly interpret some concepts, especially the model estimates vs hypothesis tests. Additionally focus on the interpretation of models instead of just the hypothesis tests will be considered for future semesters. It is unclear if the project prompt led to some confusion for future semesters. It is unclear if the project prompt led to some confusion of the task at hand and confusion of methods employed but the project descriptions will be carefully considered to try to make this more clear and hopefully more consistent across the students.
Comments/ideas for improving the assessment process

In future assessment with large(r) multi-section courses, it is recommended that "signature" questions on an exam be devised that directly relate to the learning outcomes despite the increase in work required to digitize and store exam responses. Using projects involves quite a bit of inherent variability in the ways students may or may not be meeting the outcomes and may require more time to complete the assessment.

 none None, as the project provided a good platform to assess student ability to interpret model estimates and tests. But this is a difficult concept in a complicated model and some struggled to do this successfully.

 

PDF of STAT 217 Q-Core Assessment Report