Monday, June 4, 2012

Timed Writing: 6/3/2012


Time: 10 minutes
Prompt: "... features, well formed by nature, modelled by art..."
Source: 6 x 5 x 6 + 3 x 6 + 5 = 203 = Ivanhoe by Sir Walter Scott

Someone -- one of the early and/or great statisticians, or maybe just a popular professor -- said that in statistics, as in art, one should avoid falling in love with one's model. The first time Vince heard the quote he laughed. Later, when he'd spent some time modeling multivariate time trends using mixed effects regression models, he understood that it must not have really been a joke -- not unless maybe it had been misquoted and misattributed, and it was really an artist joke. But then it didn't make sense that it had been told in a statistics class. He put it from his mind.

Early on, Vince avoided talking to non-statisticians about his work because they wouldn't understand and the entire conversation would be a waste. He would have to walk the listener through the most elementary concepts just so that he or she would possess enough of the necessary vocabulary to comprehend his heavily watered summary of the theories behind his pursuits. Later, he avoided talking to people for another reason: conversation was a sure way to squander time better spent with maximum likelihoods estimators, multiple imputation algorithms, or validation methods for measurement error corrections.

One day, an aesthetically pleasing female student in his program asked Vince to attend a Christmas party with her that evening. Vince never even looked away from his monitor. He sent her an email listing half a dozen dilemas that he was busy untangling, all of of which would prevent him from joining her. He wasn't rude. It was just that he could type faster than he could talk, so a reply by email was a more efficient use of his time. She received the email on her phone and read it while she was still standing beside his desk, waiting for him to acknowledge her presence. She didn't say anything. She just read the message, looked over his shoulder for a few seconds while he worked, and then walked the seven feet to her own cubicle and sat down.

Two hours later, Vince received a reply to his email reading, "see attached." The text file consisted of some ninety-five lines of R code referring to the data set on which he was laboring. He ran the code to discover that the girl had provided solutions to all of the problems that he had catalogued in his initial message. Among the command annotations, random text phrases were scattered. Some of them were snatches of poetry. They had nothing to do with the statistical analysis, but they did share a common theme: the benefits of social engagement. Vince deleted the irrelevant characters from the page -- cleaned up the code.

Vince did not not resent the girl's help. He appreciated it greatly. It allowed him to proceed to the next stage of his analysis all the sooner.  He meant to thank her. Later, he couldn't remember whether or not he ever had. He did not attend the Christmas party. He was vaguely aware of movement in the shared office area when the girl and some of the others exited. His mind wandered for a moment. Why would such a talented mind as hers be distracted by a social engagement. The question bothered him for only that one moment and then he returned to his work.


(about my timed writing exercises)


Creative Commons License
This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivs 3.0 United States License.

No comments: