Importance of communication and how to improve it

  • Thread starter Thread starter Grelbr42
  • Start date Start date
  • Tags Tags
    Communication
AI Thread Summary
Improving communication, especially in technical discussions, is crucial for effective collaboration. A recent experience highlighted the challenges of extracting detailed information from a coworker about a PERL script used to calculate uncertainties in data analysis. The initial inquiry about how uncertainties were produced led to vague responses, emphasizing the need for clearer, more specific questions. Instead of asking about the script directly, focusing on the underlying methodology or the algorithm could yield better insights. The importance of documentation was also underscored, as the absence of comments and clear variable names in the code complicated understanding. Suggestions for better communication included asking for flowcharts of the code and ensuring that the code is well-commented to facilitate easier discussions in the future.
Grelbr42
Messages
68
Reaction score
125
So the generic question here is, how to improve communication?

Recently I started a review task for a coworker's analysis. One of the items in the report for the analysis is a graph with a linear fit of some data. And the linear fit has uncertainty reported in the fitting parameters. So I ask "What did you do to produce these uncertainties?" His response is "I calculated them." Um... Click click click. "Yes, but what did you *do*?" So he starts to show me his PERL script. "Wait. OK. What does the PERL script do?" He tries to show me more script. "No no. Source code is not documentation." He looks quite upset with me but pauses. "What does the script do to calculate the uncertainties?" Suddenly he's back in the groove, walking me through 200 lines of PERL. And suddenly I see that he has included a covariance matrix. Sigh. I know what the client will do with *that*. They have a very large pile of such things behind their facility. On the far side of the parking lot behind the trees. "So you used the textbook formulas, fed the data in, and got reports of the uncertainty of the fitting parameters, yes?" Well... Yes. But my PERL script... "No no, that's quite alright. I have what I needed."

How could I have asked this question so that my meaning got through?
 
Physics news on Phys.org
If this is the first interaction of this sort, then I think it is natural to have some iteration on questions to get to the heart of the matter. I suppose in this instance you could have asked what type of uncertainty or equations are implemented by the code. You kept asking what he did, and he definitely wrote code. You also asked what does the script do. These are all easily interpreted to require answer that talk about the script.
 
Grelbr42 said:
"No no. Source code is not documentation."
Was the code commented? Was there a good summary of the code in the comments at the top of the file?
 
Grelbr42 said:
How could I have asked this question so that my meaning got through?
Be careful. You could find yourself with the added responsibility of supervising this person. Extra work, much frustration, and at the same pay.

Some suggestions:

What is the algorithm implemented in your code?

Sketch the flow chart of the your code.

Sketch the flow chart of the algorithm in your code. Yes. Now. Freehand.
 
berkeman said:
Was the code commented? Was there a good summary of the code in the comments at the top of the file?
No comments whatever. Variable names like xj and yq.
 
How could I have asked this question so that my meaning got through?

"What underlying methodology did you use to base your calculations on?"
 
Just ONCE, I wanted to see a post titled Status Update that was not a blatant, annoying spam post by a new member. So here it is. Today was a good day here in Northern Wisconsin. Fall colors are here, no mosquitos, no deer flies, and mild temperature, so my morning run was unusually nice. Only two meetings today, and both went well. The deer that was road killed just down the road two weeks ago is now fully decomposed, so no more smell. Somebody has a spike buck skull for their...
Thread 'RIP George F. Smoot III (1945-2025)'
https://en.wikipedia.org/wiki/George_Smoot https://physics.berkeley.edu/people/faculty/george-smoot-iii https://apc.u-paris.fr/fr/memory-george-fitzgerald-smoot-iii https://elements.lbl.gov/news/honoring-the-legacy-of-george-smoot/ https://www.nobelprize.org/prizes/physics/2006/smoot/facts/ https://www.aps.org/publications/apsnews/200611/nobel.cfm https://inspirehep.net/authors/988263 Structure in the COBE Differential Microwave Radiometer First-Year Maps (Astrophysical Journal...

Similar threads

Replies
4
Views
1K
Replies
4
Views
2K
Replies
30
Views
3K
Replies
5
Views
2K
Replies
1
Views
2K
Replies
6
Views
3K
Back
Top