This is an interesting and thought provoking piece, about language, algorithms, and all the nasty isms (sexism, racism….). Some researchers at Princeton used a machine learning algorithm to read 800 Billion words online (original article) and then looked at the ties between words. So words like management and salary are tied more to men, words like family and home are tied more directly to women. In a sense the findings aren’t extraordinary, it has been a part of known or accepted knowledge in psychological or cultural studies for years. In another sense, it is extraordinary in that a computer with an, almost, inordinate amount of data figured out that most human speech is biased. Its one thing for someone from any group to say I feel like I experienced discrimination, usually it is shrugged off as paranoia, attention seeking, over reactions or numerous other “excuses”. But for a computer to point this out supports those people who have felt like they experienced some form of discrimination. Knowing it is important, having evidence to support the knowledge is equally important, making changes based on this is the most important.
I know I’ve read this somewhere, but cannot remember where, that the best research is that which makes you say, everyone knows that. But as a researcher, most times people don’t know that, which is why we did the research. Second and third studies don’t get published. So this article from Bloomberg talks about how a slump in people attending college will hurt the future productivity of the United States. Sounds like everyone should know that, and perhaps they do. The real question that is on the minds of every administrator at every one of the 4500 colleges and universities in the US is, how do we change it? Well perhaps a startup will help? Viridis Learning wants to work with community colleges to create a “passport system” to link business needs with “skills and competencies” the school can teach. It sounds like a good idea, potential employers can identify needs, and schools supply candidates who can meet those needs. If our current student decline and future productivity loss can be stemmed by a company like this, it sounds like it will fulfill a need. My guess, though, is the devil is in the details. Don’t students already receive certificates, diplomas, and transcripts which sound a lot like a passport? Who then will decide what classes are needed? I’ve heard arguments pro and con that students in x field don’t need to learn public speaking, or research paper writing, or x level of math. Perhaps this startup will add data to the argument if employers do decide, yes we’d like our employees to be able competent at y level of public speaking, writing, or math.
This article discusses a recent survey of faculty knowledge and use of OER (Open Educational Resources). From my own experience with other faculty I would agree with the findings. Around five percent of faculty in higher ed use OERs. Why don’t they use more? The article gives the top 3 reasons, the top 2 are: not enough resources or difficulty finding resources, both of which really boil down to time. If there aren’t enough, sure you could save students money but it takes time to come up with those other resources. You can’t find them, it would take more time to find them, or more time is needed to create them yourself. I think these are valid reasons, who has extra time? I’ve tried to use OERs in the past for history and there would often be some period of history, not yet covered, not available. Its much easier to rely on the textbook which is all there, prepackaged. But I think the discussions of OERs needs to be reframed a bit, not from the standpoint of looking “out there” for full resources to replace what we’re using, but looking at what we already have. Most faculty I know already have full notepacks on their subject. And as I remember 25 years ago, I too bought these notebooks, and uh, ahem, probably used them more than the textbooks. If we already have our own resources, the step from using those and dropping a textbook is a much smaller step, and takes much less time. Finding the extra resources we would need to fill in holes in our notes is a much easier task to accomplish using existing OERs, than using our notepacks and expecting OERs to replace our books.
The final piece I looked at today, I think touches on reframing or rethinking extant systems. The author gives an excellent example of how the whole business model of newspapers would make no sense if it was floated in a boardroom today, but because we’ve always had newspapers it makes sense to keep them. I think it is a very interesting point of view for considering many of the models and businesses we have in place today: tv, radio, newspapers, textbook companies. Are we only using existing systems because we’ve always used them, or because they make sense today?