Last week, I was cautiously optimistic about New England’s chance of pulling out another Super Bowl victory and was scrambling to get the issue out earlier than kick-off. I was mildly encouraged by the coincidence of it being issue 87, corresponding to Gronk’s uniform number. Still, I was not as successful as I would have liked to be. Even though I sent it out before the game, it likely was not early enough. So, many of you might passed it over without much notice. Nevertheless, as New Englanders bask in the dying embers of victory’s glow, here is another issue.
This week has a pretty interesting mixed bag of items. The first couple of links are wonderfully inventive visual representations of information more than articles. The second pair of links are articles more typical of what I usually share, although they are more about implications of decisions that are made for us with some remarkable consequences.
This week’s “If you read only one article…” is the last one. The fact that a study highlights how students’ standardized test scores were markedly different when transitioning to a computer-based system is interesting enough. I am as interested by this information being shared about a test almost four years ago with almost no information more current than that. Yet, the first item below is super cool.
Hope everyone enjoys the week and the coming vacation.
Here are three+ curated articles about education, technology, and evolutions in teaching.
This one is for all the science teachers out there that subscribe to the newsletter. It is also pretty cool for English teachers too. It is an interactive web-based periodic table that renders an element inspired haiku for ever element on the table. How cool is that? Do not let the 11-minute read time frame fool you, it is only because there are 119 elements. That is an impressive number of haikus.
If any chemistry teacher is not as impressed with the poetic infusion, take look at Science magazine’s most recent visual representation of the periodic table, a historical virtual tour of the table.
Giving algorithms a sense of uncertainty could make them more ethical – MIT Technology Review – Karen Hao (4-minute read)
I found this article both interesting and infuriating. On a fundamental level, the fact that there is an acknowledgment of what algorithms are actually designed to do and how they are being used incongruent ways is strong. Even contrasting how algorithms are designed to pursue a single mathematical objective with the incompatibility of humans wrestling with multiple desires highlighted an interesting dilemma. Yet, the very framing of the problem becomes a fundamental paradox.
The article proceeds with a flawed notion. If only we introduce uncertainty maybe the algorithms will be better. Adding multiple options, percentages, or tendencies could potentially increase the usefulness of algorithms. This is where the article is headed. The first source explains, “There are many high-stakes situations where it’s actually inappropriate — perhaps dangerous — to program in a single objective function that tries to describe your ethics.” Then continues to advance possible techniques for expressing the idea mathematically as a way to force a tool to function in a way that defies its function.
Maybe, just maybe, we should not use algorithms to make ethical decisions. I am not really sure why that never seems to gain any purchase.
Scores Were Lower When Mass. Students Took PARCC Exams on Computers, Study Finds – EdWeek’s Digital Education blog – Benjamin Herold (5-minute read)
My first reaction to seeing this headline was “No fooling?” I have to admit that I was not familiar with the term “mode effect” or at least knew what it was called but just about every English teacher I know was deeply concerned with the transition to computer-based test. While some might quickly dismiss that as aversion ot change. The concerns were far more nuanced than that, not that any decision-makers in power were likely to listen.
The main concern was that so many reading strategies taught in the classroom revolve around marking up texts, not to mention the prospects of pre-writing strategies involved in composing written responses. A computer-based test has a digital interface mediating the freedom of those activities and potentially interfering significantly. Thus, it is not shocking at all that there was a significant difference in the “mode effect” between the math and ELA tests.
What is particularly interesting about this article, however, is how there is hard data in the form of statistics to highlight the negative effects but none to validate the claim “By 2018, there was no significant effect.” I guess we are just supposed to take the word of the spokesperson for the outfit that took over the test items from the defunct PARCC. This strikes me as one more reason to rethink the value of these tests anyway, regardless of Massachusetts’ “hold harmless” policy. Certainly, the promised advantages of computer-based tests have not really come to pass. The scoring is not faster and the tests are more expensive than ever. Lastly, all of this information is based on the results from a 2015 test which only reinforces my claim that by the time anyone is even looking at standardized test scores they are looking at a picture of what may have been happening two to three years ago if they offer any real insight at all.