We began our class on experiments and quasi-experiments with some experimental music.
Nothing too weird, just the Dave Brubeck Quartet's classic "Take Five,"
from the experimental 1959 album Time Out.
What makes it an experiment? They weren't recording in lab coats or
anything like that, but as with the experiments we discussed in class,
they manipulated some variables -- and kept others constant -- to see what
would happen. In this case, Brubeck and his bandmates composed songs in
unusual time signatures, departing from the traditional
one-TWO-three-FOUR count of most western popular music. (Try counting
along to "Take Five"; you'll get the title right away...) However,
Brubeck and company retained the traditional harmonic and melodic
structures of jazz, as well as the traditional instrumentation of a
quartet. They were asking, in musical form, the basic question that
drives all experimentation: what happens if we change X?
For this week's blogging question, let's either imagine or describe an information-related experiment of some kind. It could be an experiment in the strict sense with laboratory conditions, or a quasi-experiment that adapts traditional scientific principles to less controlled settings, or something at the edges of traditional research, in the spirit of the "Take Five" example above. (Remember what I said in class about an intelligent hoax being halfway to a social experiment if conducted the right way.) You could use a real-world example, or you could describe an experiment you'd like to conduct -- within the boundaries of legality and ethics, of course.
To keep us grounded, your example should have some connection to information research -- which, as we've been discussing, can be construed broadly, and includes approaches from the social sciences, the humanities, the arts, design, education, and even music. Also, to keep us focused on the specific principles of experimentation, please describe how the three kinds of experimental variables (independent, dependent, and controlled) manifest themselves in your example. That might be a straightforward answer for some examples, or you might have to think creatively if your example is something experimental rather than a strict experiment.
For example, in my musical example above, the independent variable is clearly the choice of time signature -- that is, the choice to use 5/4 and not the normal 4/4. In other words, it's the thing that Brubeck and his bandmates deliberately changed. The dependent variable would be the band's performance: could they write a coherent song under this constraint, and how might it sound? As it turned out, they managed to write a classic tune with an infectious groove, and proved that it could be done in an odd time signature. As for controls in this example... I'd have to think more about it. (Maybe the choice to retain traditional harmonic structures is one of the controls? The chord progression of the song isn't actually weird or experimental.) Anyway, that kind of thinking about what constitutes an experiment is the point of this blog question -- and feel free to write about a much more conventional example if you prefer!
Lecture slides and follow up to our last class will appear here later today -- I wanted to get the blog question posted first.
For this week's blogging question, let's either imagine or describe an information-related experiment of some kind. It could be an experiment in the strict sense with laboratory conditions, or a quasi-experiment that adapts traditional scientific principles to less controlled settings, or something at the edges of traditional research, in the spirit of the "Take Five" example above. (Remember what I said in class about an intelligent hoax being halfway to a social experiment if conducted the right way.) You could use a real-world example, or you could describe an experiment you'd like to conduct -- within the boundaries of legality and ethics, of course.
To keep us grounded, your example should have some connection to information research -- which, as we've been discussing, can be construed broadly, and includes approaches from the social sciences, the humanities, the arts, design, education, and even music. Also, to keep us focused on the specific principles of experimentation, please describe how the three kinds of experimental variables (independent, dependent, and controlled) manifest themselves in your example. That might be a straightforward answer for some examples, or you might have to think creatively if your example is something experimental rather than a strict experiment.
For example, in my musical example above, the independent variable is clearly the choice of time signature -- that is, the choice to use 5/4 and not the normal 4/4. In other words, it's the thing that Brubeck and his bandmates deliberately changed. The dependent variable would be the band's performance: could they write a coherent song under this constraint, and how might it sound? As it turned out, they managed to write a classic tune with an infectious groove, and proved that it could be done in an odd time signature. As for controls in this example... I'd have to think more about it. (Maybe the choice to retain traditional harmonic structures is one of the controls? The chord progression of the song isn't actually weird or experimental.) Anyway, that kind of thinking about what constitutes an experiment is the point of this blog question -- and feel free to write about a much more conventional example if you prefer!
Lecture slides and follow up to our last class will appear here later today -- I wanted to get the blog question posted first.
