Sunday, March 27, 2011

Efficient Multi-tasking – The Myth that does not seem to go away

The other day at work I walked to someone’s office for a pre-appointed meeting. I was on time, warmly welcomed and gestured to take a seat. Soon enough I realized that the speaker-phone on the desk was on and there was a live call that my host was intermittently joining and muting. Regardless, our discussion slowly started. Soon, I noticed that he was also checking his blackberry calendar from time to time. Our meeting continued with an occasional spaced-out look from my host as he tried to keep an ear to the ongoing call. Some times, he made a few disjointed comments to the unseen participants of the call, efficiently muting it back. Suddenly, distracted by some supposedly “hot” email chime, he began texting to someone. At this point, I had it. I politely suggested that I might have come at a wrong time and perhaps we could reschedule. “Oh no, no…”, he was superbly gracious. “Your time is important, so let’s continue and finish the conversation now. Don’t worry about the call or the SMS. I am a pretty good multi-tasker”….

It is time to seriously challenge the “coolness” and glory of such multi-tasking in today’s culture. Look at the well known 2009 Stanford University study by Clifford Nass involving several self-declared super-multi-tasker youngsters. These kids were at the leading edge of simultaneous IM conversation, texting, emailing, social networking while doing everything else and claimed to have full cognitive control. However, when given a series of mental tasks like puzzles and memory games, these chronic multi-taskers consistently under-performed relative to low or non-multi-taskers. Check out this PBS interview of Dr. Nass on his research. 

So what’s at play here..? The multitude of web and information technology has made multi-tasking both easy and endemic.  But every time we switch from one task to another, there is a “switching cost” of time, as we turn off one part of the brain and turn on another. Second, human brain functions by relating data, picture or information in front of us. As we switch from task to task, images or data from all other tasks tend to clog the brain, impeding any analytical effort that needs only selective data from that mess. Multi-taskers’ rationale is that they do five things at once because they don’t have time to do them one at a time. Turns out, they might be more efficient if they actually did things one after another.

Even for a computer with a single core micro-processor, multitasking involves time-sharing with only one task active at a time. Typically, these tasks are rotated through many times a second, with a miniscule but finite ‘Context switching’, time lost in between. However, with multi-core processor computers, each core can perform a separate task simultaneously.

For those more geekily inclined, there is a nice 2001 blog piece by Joel Spolsky that gives a simple example from a programming point of view to establish that:
1. even without any task switching costs, sequential processing gets you results faster on average.
2. the longer it takes to task switch, the bigger the penalty you pay for multitasking.

To elaborate the concept, let me borrow a graph from project management paradigm (cited in  from Gerald Weinberg’s “Quality Software Management –Systems Thinkingthat conceptualizes the relative time share between a software project and the wasted time in context switching as more projects are added to the workload. Be sure to check a similar blog titled Twitter Curve by George Mulhern. 
The point, however is NOT to avoid or shun these fantastic tools that human mind is innovating. I love Twitter, Linked-in, Facebook, texting and SMS. The point is that our comfort and expertise notwithstanding, we often take a myopic view and believe that we can text, email, phone and think at the same time without sacrificing the quality of any of these. The going paradigm often is that we can IM, watch a movie and do homework or solve problems without degrading the quality of the solving process. Overwhelmingly, studies indicate that we really can’t, at least not without degradation in terms of time, quality or ability to think effectively.

Technically speaking, it is possible that after thousands (perhaps millions) of years of such multi-tasking, the human brain might evolve itself to a multi-core processor to bridge the gap of context switching between individual tasks. (Those who are not convinced of Darwinian science of course won’t even have such hopes; the only possibility for them might be to petition the Creator to go back to the drawing board and re-design the human brain for multi-core processing).

Let me end today with another blog article by Myra White  where she mentions psychologist Csikszentmihalyi’s work to underline that some of our most meaningful and creative moments happen when we’re so absorbed in the task at hand that we become one with the task and lose track of everything else. Many of the Aha or the Eureka events and innovation happen during these very moments.