Posts by linger
Last ←Newer Page 1 2 3 4 5 Older→ First
-
Hard News: Drunk Town, in reply to Tom Semmens,
These things do not have to be a simple binary choice.
Drunkenness is also common in Japan, without resulting in public violence, and in the absence of totalitarian government.
Which seems to weaken the causal connection you are trying to make. -
At least under currently available technology, solar makes more sense as an assist to (water) heating than as a main source of electricity.
(The economics are a little more persuasive in say, Arizona.)Decreasing energy use per capita by increasing the number of people doesn’t seem an ambitious target, nor easily achievable given other limits on city infrastructure, most notably sewage treatment.
Increasing the population to 15 million while preserving a functioning economy would also require that NZ find other sources of income than agriculture (e.g. it’s not likely we could have 3x as many cows without stuffing up our environment [more]).We should be aiming to decrease energy use per capita, period – and thereby decreasing NZ’s total energy and resource use.
-
Hard News: How to make more awesome use…, in reply to tarlen,
after posting, edit the post and you can add another. Then edit again to add a third.
Does this process allow photos to be anchored to a specific point in a comment? Or do all attachments just fall to the bottom of the comment?
-
Hard News: Briefing, blaming, backing down, in reply to Bart Janssen,
the analysis has created something odd.
Not really; it has created something predictable – the largest effects are identified for the factors that are most consistently associated with better learning outcomes (assuming for the moment that those were measured appropriately; that’s another can o’ worms). We should expect that the more specific factors, which have a more direct causal connection, will be more consistently associated with outcomes than class size per se. To that extent, the results are not “odd” at all.
But (and this is where the policy response was so deeply flawed) it does not follow that class size is unimportant, since many of the more specific factors listed are only easily applicable in smaller classes.The basic methodological problem is that, given the likely connections between factors (which are recognised by Hattie), a simple regression model should not have been used; a structural equation model would be more informative (though considerably more complicated). Hattie's interpretation of the results does not entirely ignore the problem, but it does suggest he is not fully aware of the fundamental limitation of his statistical model.
-
Hard News: Briefing, blaming, backing down, in reply to 81stcolumn,
most measures conflate likeability with effectiveness
though this isn't entirely unjustified, the common ground being the effect on student motivation.
-
Hard News: Briefing, blaming, backing down, in reply to merc,
built up the stock
baaaaa!
wow indeed. -
Hard News: Briefing, blaming, backing down, in reply to Ross Mason,
Actually, it’s not much of a conundrum: “class size” most likely shows a small independent effect because most of its actual effect lies in facilitating many of the higher-ranked factors (such as “feedback” and “direct instruction”). Which is to say, the listed items are not independent.
[Edit: as also pointed out by Rob Stowell above.]
Hattie’s substantive point, that smaller class sizes do not necessarily mean that teachers change their teaching styles, is possibly true – but when they do, the effect is huge. The problem is that Hattie’s methodology must statistically link any increase in performance to the more specific changes in teaching, rather than the more general decrease in size that makes these changes possible.
-
Isn't the Herald one big scheduled outrage?
-
Hard News: Sub Mission, in reply to Hebe,
Rarely, the two species can exist peaceably side by side and even intermarry.
As when subbing for the Dom ? :-P Sorry, couldn't resist.
-
I'm a pretty good editor, even of my own writing -- but I do have a considerable advantage in that I'm usually not working to a constant short deadline. Most of the time I can leave my stuff sitting around for a week before coming back to it.
The concomitant disadvantage is that often I have to leave things for a week ... or a year ... in order to do other parts of my job, not least proofing student research writing -- after which, I need at least half a day just to work out WTF I was doing.
Occasionally, things end up being left far too long. Case in point: I started editing a book 6 years ago which, all going well, will finally make it into print this year.