Comparing Absolute Deviation to Mean Absolute Deviation

Click For Summary

Discussion Overview

The discussion revolves around measuring variability in sensor data using absolute deviation and mean absolute deviation. Participants explore the validity of comparing these two measures and their implications for assessing the precision of data points when the true value is unknown.

Discussion Character

  • Exploratory, Technical explanation, Debate/contested, Homework-related

Main Points Raised

  • One participant proposes using absolute deviation and mean absolute deviation to quantify variability in sensor data, suggesting that the percentage deviation between the two could indicate variability.
  • Another participant requests clarification on the initial description and asks for an example to better understand the approach being taken.
  • A different participant questions the validity of concluding that a data set is accurate based solely on the absolute deviations being within ±5% of the mean absolute deviation.
  • One participant seeks further clarification on what is meant by "measuring the precision" and what specific arithmetic operations are involved in the analysis.
  • A later reply provides context by explaining that the participant is involved in nuclear counting for medical imaging and is assessing the consistency of counts as a part is changed.

Areas of Agreement / Disagreement

Participants express varying levels of clarity and understanding regarding the initial proposal, with some seeking more detailed explanations and examples. There is no consensus on the validity of using the proposed method for assessing accuracy.

Contextual Notes

Limitations include the lack of a clear definition of "measuring precision" and the absence of specific arithmetic steps outlined in the discussion. The context of nuclear counting for medical imaging adds complexity to the analysis.

Nyasha
Messages
127
Reaction score
0
Hi Guys,

I am trying to measure variability in a part for my sensor which l do not know the true value. So l decided that a good way to measure variability in this case would be to measure precision of my data points as l change this part on the sensor. So l was wondering, can l compare the absolute deviation and mean absolute deviation and calculate the percentage deviation between the two, and then use the percentage deviation as a number which tells me something about variability in this part.


Thanks.
 
Physics news on Phys.org
Your description is not very clear. Please flesh it out more, perhaps with an example.
 
haruspex said:
Your description is not very clear. Please flesh it out more, perhaps with an example.

I calculated 15 absolute deviations and their mean absolute deviation. Would it be valid for me to say my data set is accurate if all the 15 absolute deviations are within ±5% of the mean absolute deviation ?
 
No, I meant a much more thorough description of what you are doing. What do you mean by "measuring the precision"? What does this involve? What, arithmetically, do you then do with the observations?
 
I am doing nuclear counting for medical imaging and l do not know the 'true' estimate of counts for my system. So to see if my system is operating consistently as l change a certain part, l am going to measure its precision. That is, as l change that part, are my number of counts within the same range.
 

Similar threads

  • · Replies 42 ·
2
Replies
42
Views
6K
  • · Replies 4 ·
Replies
4
Views
3K
  • · Replies 5 ·
Replies
5
Views
2K
  • · Replies 11 ·
Replies
11
Views
8K
  • · Replies 3 ·
Replies
3
Views
3K
Replies
4
Views
1K
  • · Replies 4 ·
Replies
4
Views
4K
Replies
2
Views
2K
  • · Replies 6 ·
Replies
6
Views
4K
  • · Replies 2 ·
Replies
2
Views
2K