I don't know where the paper can be freely accessed. Looking at the last sentence of the abstract, I wonder if the real answer is to bring back ability grouping for all students.

http://www.nber.org/papers/w20453
Does Gifted Education Work? For Which Students?
David Card, Laura Giuliano
NBER Working Paper No. 20453
Issued in September 2014
Abstract:
Education policy makers have struggled for decades with the question
of how to best serve high ability K‐12 students. As in the debate
over selective college admissions, a key issue is targeting. Should
gifted and talented programs be allocated on the basis of cognitive
ability, or a broader combination of ability and achievement? Should
there be a single admission threshold, or a lower bar for
disadvantaged students? We use data from a large urban school
district to study the impacts of assignment to separate gifted
classrooms on three distinct groups of fourth grade students:
non-disadvantaged students with IQ scores ≥130; subsidized lunch
participants and English language learners with IQ scores ≥116; and
students who miss the IQ thresholds but scored highest among their
school/grade cohort in state-wide achievement tests in the previous
year. Regression discontinuity estimates based on the IQ thresholds
for the first two groups show no effects on reading or math
achievement at the end of fourth grade. In contrast, estimates based
on test score ranks for the third group show significant gains in
reading and math, concentrated among lower-income and black and
Hispanic students. The math gains persist to fifth grade and are
also reflected in fifth grade science scores. Our findings suggest
that a separate classroom environment is more effective for students
selected on past achievement - particularly disadvantaged students
who are often excluded from gifted and talented programs.

The fact that this paper is behind a pay wall prompted me ask the following quesion.

I often see this at 'pay walls' for research papers:-


We have a collection or highly gifted people on this site, some of whom are no doubt very well positioned in US society. Why can a federal employee access this but not a US tax paying US resident?

What is the rational behind this?

This question does have relevance to GT education because I find a lot of useful research papers just out of reach due to pay walls with the above blurb.

An ordinary taxpaying US resident CAN acquire a copy, for $5. The rationale is that the anti-tax cohort have, in many cases, succeeded in substituting point of sale models financing/supplementing select government functions, where that's possible.

Free here:
http://economics.virginia.edu/sites/economics.virginia.edu/files/applied/giuliano.pdf

Some may conclude that those who are selected into a program based on testing well, continue to test well.

From reading through the study (preliminary draft dated March 2014, hat tip to Zen for the link) some may say that the "treatment" offered in the study was more in tune with the needs of the cohort of high achievers of average IQ, hence the greater gains in achievement for that cohort of pupils.

This may be a valuable finding with several practical applications for education, including:
1) Encouragement for replicating the classroom practices from the study, so that more schools support the ongoing academic achievement for high-achieving students of average IQ. Many general education programs may not currently be meeting the needs of high-achieving students of average IQ, therefore general education programs may benefit from these reforms.
2) Underscoring the need to design/develop/implement classroom practices which acknowledge and serve the needs of gifted students, in order to support gains in academic achievement for this overlooked, underserved group.

These findings may be different than concluding that gifted education programs ought to target, select, and serve primarily high-achieving students of average IQ, based on observing that this cohort has benefitted the most from the classroom practices in this study.

Okay, read Most of the study.
First off, they state that all students cover the same curriculum and take the same statewide tests.
This is a pet peeve of mine: it has been established again and again that if you want a study to show that congregated classrooms lead to higher achievement gains for high ability students, teach them an accelerated curriculum and test them on that. If you want to show that mixed ability classrooms don't harm them, teach a standard curriculum and test that.
If the kid is nit challenged or engaged by the curriculum, it doesn't matter for their achievement what kind of classroom they are in (student satisfaction may be higher, though).
Further down, my knowledge of statistics isn't good enough to understand most of what they've been doing, but these numbers scream at me loudly and clearly:
There was a gifted classroom established in every school where at least one child was identified as gifted for that year. Usually there were no more than three or four. After which the classrooms were filled with another 15 or 20 of the non-gifted top achievers, whose achievement scores, naturally, were higher than predicted by their ability scores.
I fail to see why that would be called a gifted program at all - teaching a standard curriculum to a classroom full of above average motivated high achievers, with a couple gifties thrown in - of course this is a program that's perfect for the high achievers, but why would gifties thrive?

Edited to add that it appears (I only skimmed that section) that whereas high achievers moved from classrooms with average peers to strong peers. hat placed gifted kids moved from classrooms with strong peers achievement wise (due to being found disproportionately in high SES schools) with (possibly) some individual attention into classrooms with somewhat stronger peers with (presumably) no more individual attention.

So moving from one unsuitable environment to another. Why would that not be helpful...

If you take a non-disadvantaged student with IQ GE 130, enrolled in 4th grade, and give them a standardized test on 4th-grade material, then you can expect them to score in the 95th percentile or higher when no gifted services are being offered. If offered, there isn't much growth to be gained, according to this particular measure. So, where are the gains to be noted for this cohort?

The gains are to be noted when age is factored in, and you find that the student is 7 or 8 years old. The gains are to be noted when tested on material at 5th-grade level or higher.

I think that the authors of this paper have a very Lilliputian view of how high achievement can get with gifted people, especially gifted kids.

It ought to be obvious to the authors ( but clearly isn't ), that if the ceiling is too low and the starting value is close to it ( pretty well at it given margins of error and EF at that age) then one isn't going to observe a large positive delta because there just isn't the head room.

A summary of this paper is now available:

Who Benefits from Education for the Gifted?

Perhaps I shouldn't have but I stopped reading and dismissed the whole thing as anti-gifted propaganda once I noticed that the y axes were quite drasticly misaligned between the left and the right graphs...