Which Value to Use: % Discrepancy or Standard Deviation?

Click For Summary
SUMMARY

The discussion centers on the choice between using percentage discrepancy and standard deviation for analyzing experimental data from a light interference experiment. The user reported a percentage discrepancy of around 7 and noted that the difference between the theoretical and experimental values is less than one standard deviation. A recommended approach is to perform normal QQ plots on both the raw data and the log-transformed data to determine which dataset exhibits a more normal distribution, guiding the decision on which value to utilize.

PREREQUISITES
  • Understanding of experimental data analysis
  • Familiarity with normal QQ plots
  • Knowledge of log transformation techniques
  • Concepts of percentage discrepancy and standard deviation
NEXT STEPS
  • Learn how to perform normal QQ plots using statistical software like R or Python's SciPy library
  • Study the principles of log transformation in data analysis
  • Explore the implications of using percentage discrepancy versus standard deviation in experimental reporting
  • Investigate additional statistical methods for assessing data normality
USEFUL FOR

Students and researchers in physics or experimental sciences, statisticians analyzing experimental data, and anyone seeking to improve their data analysis skills.

August Lee
Messages
1
Reaction score
0
I did an experiment about interference of light when it passes two slits. I got datas and made chart, but I'm confused which result value I need to use whether % discrepancy or standard deviation. My data's & discrepancy is around 7, and difference between theoretical value and experiment value is less than one standard deviation. How can I state about this experiment? I know their concept but I cannot utilize them :(

Thanks in advance!
 
Science news on Phys.org
That is largely a judgement call. I usually do a normal qq plot on the data and another on the log of the data (essentially the percent discrepancy). If the log transformed data is much more normal then I will use it, but if they are about the same then I will use the untransformed data.
 

Similar threads

Undergrad Error bars in Excel (standard deviation, standard error, percentage)
  • · Replies 4 ·
Replies
4
Views
3K
Standard deviation and standard error
  • · Replies 1 ·
Replies
1
Views
2K
Undergrad Minimum deviation in prism spectroscope
  • · Replies 1 ·
Replies
1
Views
2K
Undergrad Standard deviation question -- population std vs sample std
  • · Replies 5 ·
Replies
5
Views
3K
Undergrad Error propagation and standard deviation
  • · Replies 3 ·
Replies
3
Views
7K
Calculations using Standard Deviation and Mean
  • · Replies 2 ·
Replies
2
Views
2K
Graduate Using standard deviation values as independent variables
  • · Replies 1 ·
Replies
1
Views
2K
Understanding the meaning of "expected fraction" (Statistics)
  • · Replies 4 ·
Replies
4
Views
2K
High School Standard Deviation as Function of Sample Size
  • · Replies 24 ·
Replies
24
Views
7K
Undergrad How to implement proper error estimation using MC
  • · Replies 3 ·
Replies
3
Views
3K