Anything But Random

When good research goes bogus

Suzanne Tamang — Fri, 13 Jan 2012 19:36:36 +0000

The news about Dr. Dipak Das and the 145 cases of falsified data is devastating to me. I was aware of the recent flurry of retractions in the scientific literature, and was even discussing the misconduct of that Harvard professor as recently as Monday with my too smart friend. But something about these accusation, and his response, is much more alarming.

Despite over 150 articles in PubMed on cardiac health, and his role as the director of the U Conn’s Cardiac Research Center, his major contribution will now be to what is becoming all too familiar — bogus science. The major issue noted by the 3 year review team involves image manipulation of study data. It gets better, they aren’t just counter-factual, the review board determined that some make “no resemblance to any legitimate experiment”. And to top that off, he claims he does not know who made the blots. Really Das?

At first I was disgusted, but somewhat sympathetic. The pressure to publish, the need to bring in grants, I can understand how initial good intentions may have gone wrong. But the claim that you don’t know who made the blots is simply ridiculous. Your lab may have duped the reviewers, editors, and grant officers that let your by, but to you also think everyone in the world is a fool?

Why? Why? Why?

12/25/2011

Suzanne Tamang — Tue, 25 Jan 2011 22:18:03 +0000

I am in the process of updating my teaching materials for my core robotics class, a course that I really feel lucky to teach. First, it isn’t your typical classroom. It’s a communal workspace with computers, sensors, motors and other neat stuff, how could it be anything less than fun? Second, I get to talk about some very weird things I never imagined existed until I looked, like this 16th century “robot”, or automaton.

When Creativity Hurts

Suzanne Tamang — Tue, 25 Jan 2011 22:02:13 +0000

Names matter! The epidemic of stupid names for some good, some bad software products is really frustrating. The new norm seems to be calling a software something that sounds like invented baby babble.

12/17/2011

Suzanne Tamang — Mon, 17 Jan 2011 05:42:44 +0000

I was recently struggling when working more with Chinese characters. I have be come aware of my blissful ignorance to the problem of sufficient multilingual support for software development packages, like Python. I like the language very much, but this stinks! After I read about some of the issues with enhanced character support for programming applications, I sincerely feel is should be a minimum standard to support Unicode. Trying to represent something like Chinese characters is an absolute pain in Python. It is really a shame the developers providing multilingual support are the big companies, like Microsoft.

1/15/2011

Suzanne Tamang — Sat, 15 Jan 2011 20:40:12 +0000

My confession is that I sometimes I feel like a contemporary Luddite, forced to trade quality interactions for a cheaper, but new and shiny one we can only look at but not examine. If this is a reason to revolt just yet, I don’t think so. I love technology, but I feel design and intent is key. I see we more frequently break what I once read as fundamental software ethic — programs should be designed to replace the work that humans do not want or cant perform — not to carry out tasks human enjoy. Wish we were more responsible with what is becoming an increasingly noisy electronic environment, that’s all.

Like any tool we create, it needs to be applied in a specific way to be useful, and result in beneficial work. Our modern technologies built around processing enormous amounts of personal data often seem to be more of a mass scale pilot-project that may or may not go wrong versus a helpful tool that is beneficial to humanity.