Asking ChatGPT to look at my student evaluations to give me advice on improving my teaching

I took the student evaluations of my course in Computational Methods taught to juniors in engineering and asked ChatGPT to summarize how I can improve my teaching.

Based on the student evaluations, here are some key points to consider for improving your teaching:

  • Adaptive Lessons: Students found the adaptive lessons to be beneficial as they encouraged independent learning and helped them overcome their fear of failure. Consider keeping or modifying this approach to continue fostering self-directed learning.
  • Course Structure: The course was praised for its structure, but some students felt that it covered too much material. Consider evaluating the course content to determine if any sections can be condensed or removed to enhance learning outcomes.
  • MATLAB Coding: Some students desired more time dedicated to MATLAB coding. Consider incorporating additional coding exercises or providing supplementary resources to further develop students’ programming skills.
  • Pre-Tests and Testing Materials: Students expressed a desire for more comprehensive pre-tests and supplemental testing materials. Consider providing multiple pre-tests and additional practice materials to help students prepare for exams effectively.
  • Real-Life Examples: Students appreciated when the course material was related to real-life examples. Continue incorporating practical applications to enhance student understanding and engagement.
  • Formula Sheet and Exam Format: Students suggested allowing the use of a formula sheet and including more applicable problems in exams rather than focusing solely on memorization. Consider revising exam formats and providing formula sheets to support students’ problem-solving skills.
  • Office Hours and Support: Students appreciated the availability of office hours and additional help sessions. Continue offering these resources to ensure students have ample opportunities to seek clarification and assistance.
  • Flipped Classroom Approach: While some students appreciated the flipped classroom approach, others found it redundant. Consider evaluating the effectiveness of this method and potentially modifying it based on student feedback.
  • Consistency in Interaction: Students mentioned inconsistencies in how you interacted with students inside and outside the class, raising concerns about fairness. Strive to maintain consistency in your interactions with students to ensure a fair and inclusive learning environment.
  • Computer Projects: Some students found the computer projects lengthy and confusing. Consider providing clearer instructions and support for these projects to enhance the learning experience.
  • Student Engagement: Students appreciated your efforts to engage them and create an enjoyable learning environment. Continue employing varied teaching methods, incorporating humor, and making connections with students to foster their interest and participation.
  • Feedback and Improvement: Students appreciated your dedication to improvement and willingness to listen to their opinions. Continue seeking student feedback and implementing changes accordingly to enhance the learning experience.
  • Clarity of Explanation: Some students found your explanations challenging to follow due to speaking too fast. Aim for clarity in your delivery, allowing students to absorb and understand the concepts effectively.
  • Feedback and Availability: Students appreciated your knowledge and availability outside of class. Continue providing clear feedback and being accessible to students, fostering their learning and growth.
  • Cultural Sensitivity: Be mindful of cultural sensitivity and ensure your teaching approach is respectful and unbiased toward all students.

Overall, the evaluations indicate that you are a knowledgeable and caring professor who puts effort into supporting student learning. By addressing the mentioned areas of improvement, you can further enhance the learning experience for your students.

A VBA function for Cohen’s effect size

While many social scientists tend to get excited to report p-values of less than 0.05 when comparing two groups, they tend to forget to report effect size. One can get a low p-value just by having a very large sample size but it is the effect size that tells you how much pragmatic difference there is between two groups. Although there are simple formulas out there for effect size, it is better to have a formula for two groups of unequal size. I wrote a VBA program for the Cohen’s effect size and you can use it readily.  You can download it form here.  If you are not familiar with how to enter the VBA program, here is a quick tutorial.  You just have to save the excel file as a .xlsm file.

Example of usage is
where A23:A78 has the control group numbers and B24:B67 has the experimental group numbers.

Function EffectSizeCohen(ControlGroup As Range, ExperimentalGroup As Range) As Variant
'This function finds the Cohen's effect size given the numbers from the control
'and experimental group
'ControlGroup: Excel range of numbers for control group, e.g. A120:230
'ExperimentalGroup: Excel range of numbers for experimental group, e.g. A120:230
'Cohen's effect size.
'See formula at

'Putting the ranges in a column vector. Note that excel stores it as as two dimensional matrix though
ControlData = ControlGroup.Value
ExperimentalData = ExperimentalGroup.Value

'Number of entries of the two groups
ControlLength = ControlGroup.Count
ExperimentalLength = ExperimentalGroup.Count

'Calculating the average and standard Deviation of control group
ControlSum = 0
For i = 1 To ControlLength
ControlSum = ControlSum + ControlData(i, 1)
Next i
ControlAvg = ControlSum / ControlLength

ControlVar = 0
For i = 1 To ControlLength
ControlVar = ControlVar + (ControlData(i, 1) - ControlAvg) ^ 2
Next i
ControlStd = (ControlVar / (ControlLength - 1)) ^ 0.5

'Calculating the average and standard Deviation of experimental group
For i = 1 To ExperimentalLength
ExperimentalSum = ExperimentalSum + ExperimentalData(i, 1)
Next i
ExperimentalAvg = ExperimentalSum / ExperimentalLength
ExperimentalVar = 0
For i = 1 To ExperimentalLength
ExperimentalVar = ExperimentalVar + (ExperimentalData(i, 1) - ExperimentalAvg) ^ 2
Next i
ExperimentalStd = (ExperimentalVar / (ExperimentalLength - 1)) ^ 0.5

'Calculating the Cohen's effect size
'See formula at
StdMean = (((ControlLength - 1) * (ControlStd) ^ 2 + (ExperimentalLength - 1) * (ExperimentalStd) ^ 2) / (ControlLength + ExperimentalLength)) ^ 0.5
EffectSizeCohen = (ExperimentalAvg - ControlAvg) / StdMean
End Function