Friday, August 21, 2015

What Affects Student Achievement?

Here's what one researcher has found:
Just for giggles, compare direct instruction to cooperative learning, computer-assisted learning, inquiry-based learning, individualized instruction, problem-based learning, and various and sundry other recent fads.  I need to look at what the first few on the list mean and try to employ them.

Hat tip to Joanne for the link.


momof4 said...

When considering the merits of direct instruction vs the currently-preferred "discovery", "child-centered" and other constructivist methods, the TIME factor should be included; direct instruction is much more efficient (not that I've ever seen any evidence that the ed world is aware of the concept, let alone appreciates it). Constructivist practices consume LOTS of time - even if one thinks they are effective (which I don't).

CyberChalky said...

As a Mathematics Teacher speaking to another Mathematics teacher, doesn't a measure claiming to rank all possible teaching methodologies with a single variable (without reference to any other measure of population variability) seem statistically questionable?

If you investigate further into the techniques being applied, I think you may find further strangeness - the Effect Size metric appears to be a non-standard application of Cohen's d - to the point that the validity of the output is problematic.

socalmike said...

I can't see it very well - it's pretty hard to read since it's so small. (Myabe i AM getting old) Is 'small class size' on the chart?

Ellen K said...

Note the second most affective relates to the studies of Piaget. One key element of his writings was the need for students to be presented with age appropriate expectations. It used to be that the expectations were that most students would not be fluent readers until the end of first grade. Now that expectation is pushed onto kindergartners regardless of decades of students that show most students are incapable of critical tracking and decoding until after they turn six. This pushing of education into younger ages has two sources. Parents want their kids to be prodigies. They buy all kinds of baby education programs that largely have no real impact on learning and can instill some serious aversions to learning if not handled correctly. Politicians see this as another source of control and are already demanding PreK programs across the board for all kids when 15 years ago kindergarten was a half day program tops. The parents that want this are seeking free daycare and have been brainwashed into thinking this will help their kids achieve. If that was truly the case, then wouldn't we have a bunch of kids from Headstart and similar programs racking up scholarships in academic pursuits? Please bring back Piaget. Stop treating kids like tiny adults and meet them where they are so they can learn without the anxiety riddled program we have now.

Darren said...

SoCalMike: Class size is 30-somethingth from the bottom.

CyberChalky: The conclusion isn't necessarily problematic. Statisticians often "account" or "control" for the presence of other variables.

CyberChalky said...


Statisticians *can* control for the presence of other variables. That does not demonstrate that Hattie *has* controlled for other variables. In fact, there are ongoing concerns about his statistical skill (and his probity).

Perhaps this link might indicate why I remain concerned: