Palm Pilots and Swab-N-Dab
Jan. 19th, 2015 03:50 pm![[personal profile]](https://www.dreamwidth.org/img/silk/identity/user.png){kind=small}
rebeccmeisterWhile observing the somewhat-chaotic planning processes that happen in my current household, ![[livejournal.com profile]](https://www.dreamwidth.org/img/external/lj-userinfo.gif)
scrottie took to remarking that we should all get Palm Pilots to better coordinate our efforts. He's not alone in seeing some advantages to the things over the currently popular Swab-N-Dab devices such as the one I just ordered (sigh).
Aside from observing other people stylus away at their Palm Pilots back when they were super trendy, my main experience with the things occurred during a period when my PhD advisor was enamoured with them. She'd been helping with a research project to track social interactions between toddlers on a playground, where the Palm Pilots were used by a team of scientists to record different aspects of the toddlers' behavior. The data were used to construct an agent-based model of dyad formation.
By extension, she thought the Palm Pilots would prove tremendously helpful for collecting other types of behavioral data. So at her behest another grad student and I spent several hours learning how to set up Palm Pilots with drop-down menus for data collection. For much of our behavioral work, we want to know three basic things: where an individual is located (generally coarsely coded - in the foraging arena, in the nest, on the fungus garden), what that individual is doing (walking, antennating, holding still, etc.), and whether that individual is engaged in social interactions with others. With the Palm Pilot, this could be recorded with nine stylus ticks.
In practice, the interface was too clunky and slow. Our traditional method for collecting this data involved scribbling it out by hand on sheets of paper, then transcribing it into a spreadsheet. One important thing to know is there are a lot of cases where an individual's behavior didn't change much between scans. As a shorthand for this, we don't bother to rewrite everything out every scan - we just draw a line through that timeslot. Much faster.*
Our first big project with the Palm Pilots was supposed to be our bee experiments in Australia. T and I dutifully loaded the Palm Pilots into our luggage. As we discussed the details of the experimental logistics with our three other collaborators on the ground in Australia, however, the Palm Pilots quickly got a thumbs-down. Too sluggish, too many unnecessary data-processing steps. There was general agreement that the most reliable data collection method would be pen and paper (barring fire or water), but eventually we settled on a compromise: we would collect our data in teams, with one person observing the nests and relaying the behavior to a second person sitting at a computer, who would type the behavior directly into a spreadsheet. Files were backed up daily to two separate devices.
For the observations, I paired up with my friend S, while T paired with R. On our first or second watch, S and I discovered an even faster method for documenting the behavior of ~100 sweat bees in ~80 nests. After the first round, S would read back the bee behavior while I checked to see whether the bees were still doing the same thing or had changed. Then I would say yes, or no and what was actually happening. A round of behavioral observations then took around a minute, instead of five to eight minutes. And the process was so hilarious at first that it caused me to jump up and down with laughter.
I suspect that after the swab-n-dab phone arrives I'm going to continue using my paper calendar and the small notebook where I keep track of things I want to buy, scrabble scores, certain recipes (scones, crepes, lemon curd), and projects.
We shall see.
I should also note that swab-n-dabs have been useful for certain kinds of data collection, coupled with certain kinds of analyses.
Also, when I got to the lab in Texas, I discovered that here, too, there's a drawer full of abandoned Palm Pilots.
*There is a lot that a person could and should do with these datasets, including asking questions about behavior/task durations, and transition probabilities across observations. Also, we are careful about the minimal amount of time that elapses between subsequent scans; one benefit of completing a round of scans more quickly is lower overall fatigue, which ensures higher-quality scan data. Not many of the datasets have been released yet because we still have to figure out how to analyze them, write a manuscript about them, and publish them so we get some credit for our work. While we might generate one or two manuscripts through our initial effort, we have to choose our questions and analyses strategically, and leave behind many interesting unanswered questions.
scrottie took to remarking that we should all get Palm Pilots to better coordinate our efforts. He's not alone in seeing some advantages to the things over the currently popular Swab-N-Dab devices such as the one I just ordered (sigh).Aside from observing other people stylus away at their Palm Pilots back when they were super trendy, my main experience with the things occurred during a period when my PhD advisor was enamoured with them. She'd been helping with a research project to track social interactions between toddlers on a playground, where the Palm Pilots were used by a team of scientists to record different aspects of the toddlers' behavior. The data were used to construct an agent-based model of dyad formation.
By extension, she thought the Palm Pilots would prove tremendously helpful for collecting other types of behavioral data. So at her behest another grad student and I spent several hours learning how to set up Palm Pilots with drop-down menus for data collection. For much of our behavioral work, we want to know three basic things: where an individual is located (generally coarsely coded - in the foraging arena, in the nest, on the fungus garden), what that individual is doing (walking, antennating, holding still, etc.), and whether that individual is engaged in social interactions with others. With the Palm Pilot, this could be recorded with nine stylus ticks.
In practice, the interface was too clunky and slow. Our traditional method for collecting this data involved scribbling it out by hand on sheets of paper, then transcribing it into a spreadsheet. One important thing to know is there are a lot of cases where an individual's behavior didn't change much between scans. As a shorthand for this, we don't bother to rewrite everything out every scan - we just draw a line through that timeslot. Much faster.*
Our first big project with the Palm Pilots was supposed to be our bee experiments in Australia. T and I dutifully loaded the Palm Pilots into our luggage. As we discussed the details of the experimental logistics with our three other collaborators on the ground in Australia, however, the Palm Pilots quickly got a thumbs-down. Too sluggish, too many unnecessary data-processing steps. There was general agreement that the most reliable data collection method would be pen and paper (barring fire or water), but eventually we settled on a compromise: we would collect our data in teams, with one person observing the nests and relaying the behavior to a second person sitting at a computer, who would type the behavior directly into a spreadsheet. Files were backed up daily to two separate devices.
For the observations, I paired up with my friend S, while T paired with R. On our first or second watch, S and I discovered an even faster method for documenting the behavior of ~100 sweat bees in ~80 nests. After the first round, S would read back the bee behavior while I checked to see whether the bees were still doing the same thing or had changed. Then I would say yes, or no and what was actually happening. A round of behavioral observations then took around a minute, instead of five to eight minutes. And the process was so hilarious at first that it caused me to jump up and down with laughter.
I suspect that after the swab-n-dab phone arrives I'm going to continue using my paper calendar and the small notebook where I keep track of things I want to buy, scrabble scores, certain recipes (scones, crepes, lemon curd), and projects.
We shall see.
I should also note that swab-n-dabs have been useful for certain kinds of data collection, coupled with certain kinds of analyses.
Also, when I got to the lab in Texas, I discovered that here, too, there's a drawer full of abandoned Palm Pilots.
*There is a lot that a person could and should do with these datasets, including asking questions about behavior/task durations, and transition probabilities across observations. Also, we are careful about the minimal amount of time that elapses between subsequent scans; one benefit of completing a round of scans more quickly is lower overall fatigue, which ensures higher-quality scan data. Not many of the datasets have been released yet because we still have to figure out how to analyze them, write a manuscript about them, and publish them so we get some credit for our work. While we might generate one or two manuscripts through our initial effort, we have to choose our questions and analyses strategically, and leave behind many interesting unanswered questions.
