No Child Left Behind

Some time ago, around the time hat Congress passed No Child Left Behind, a reporter asked me what I thought its effect would be. I said I didn’t think it would have any.

Judging from the NAEP long-term trend scores for 17-year olds, I was right. They have hardly changed in more than 40 years. Naturally that’s what you would want to look at, rather than the scores for 4th graders or 8th graders. If they go up while graduates stagnate, it doesn’t matter. Only an idiot could get excited over that kind of sterile improvement – but since nearly everyone involved with this question is an idiot, they go on and on about it.

Reminds me of the way in which the Soviets kept increasing the production of steel, rather than steel products…

Advertisements
This entry was posted in Uncategorized. Bookmark the permalink.

84 Responses to No Child Left Behind

  1. MawBTS says:

    What do you think of charter schools/voucher programs? Seems to be topical at the moment.

    • DdR says:

      He’ll likely say that they won’t make any difference b/c it’s about the IQ of the student body and not public vs. private. Hence why CUNY was considered one of the best colleges in the country until the Ivies began allowing in Jews. Now CUNY is a functional community college.

    • gcochran9 says:

      You might save some money, if you hired non-union. If your charter schools were outside the laws of the United States, something like Gitmo, you might be able to impose discipline. I doubt if they have much educational effect if you take into account the characteristics of the incoming students, as any sane person would do.

      • Jim says:

        I suspect saving money is one of the real driving forces behind charters given the dire fiscal state of some school districts.

        • engleberg says:

          ‘Charter schools writ large are rife with mechanisms that screen out the hardest-to-educate students.’ Freddie deBoer, not a charter school fan. But it’s the great selling point for charter schools. We Expel Thugs. Public schools stopped doing this, private schools bloomed.

    • hronrade says:

      What if the charter schools were also cyborg modification shops and you added bionic arms and legs and brain segments to the students? You could turn the remedial student population into an army of super soldiers…

      • Greying Wanderer says:

        i wonder what the average IQ of SEAL types is

        • I’m guessing much higher than most people think.

        • R. says:

          I’d hazard a guess and say 115 or so.

          You could try asking at say, weaponsman.com

        • egregious philbin says:

          SEAL applicant minimum ASVAB scores (applicants must also pass physical & mental toughness screen, & meet other requirements)

          ASVAB scores are percentile ranks – oddly, the Navy is happy to add percentile ranks – not the most psychometrically sophisticated thing to do, but, whatever:) Note: both Mechanical Comprehension (MC) & Electronics Information (EI) correlate highly with Spatial Ability.

          General Science (GS) + MC + EI ≥ 170
          OR
          Verbal Expression (VE; an average of Word Knowledge + Paragraph Comprehension) + Math Knowledge (MK) + MC + Coding Speed (CS) ≥ 220

          Also, VE + Arithmetic Reasoning (AR) ≥ 110 with minimum MC = 50

          On the officer test, the Navy requires AFQT score ≥ 35. Most successful folks have AFQT scores ≥78, tho.

        • TWS says:

          Of the one I knew well he was the top IT guy for our entire facility. He was great unlike most IT guys.

    • pyrrhus says:

      As someone who has taught in grammar school, there could be three major benefits. First, excluding or expelling any kids who are disruptive (this is major),second, excluding kids who are not able to do the work, which will probably relate to the first, and third, if you are doing 1 and 2, attracting a much better class of teacher. That’s why the elite private schools deliver a better education…If the voucher scheme doesn’t allow you to do 1 and 2, it won’t have much effect, although the fact that parents of disruptive kids tend to be dysfunctional probably would result in some screening effect anyway.

      • Jim says:

        Charters might accomplish these things but it’s not what their touted for which is that they will “close the gap”.

      • Patrick L. Boyle says:

        It seems to me that one of the most effective tools for the education of children has always been ‘leaving behind’ many children. If you leave behind the disruptive, the violent, and the stupid you will be left with a better classroom. No?

        The whole idea of a program that attempts to educate everyone strikes me as an exercise in hubris. It implies that we the public educators can make successful graduates from any sort of kid no matter how defective. Why would anyone imagine such a thing?

        If that were true we should abolish the SAT, ACT and the GRE. We should do away with all entrance requirements altogether. Let’s send everybody to MIT or Cal Tech. If they can’t do long division, don’t worry, we’ll still teach them calculus.

        The logical consequence of ‘No Child Left Behind’ is that anyone can be taught anything. I except myself from this truth. I used to teach calculus and their certainly were students whom I couldn’t reach. Must have been my fault.

        A while back I dated a woman who taught Special Ed. As she explained it to me her job was more or less to terrorize the classrooms of normal kids so as to make room (physical and psychological) for her handicapped students. She didn’t care that one of her ‘special’ students’ retarded the education of a whole classroom of normal kids. She had her job to do and damn the cost.

        Her students should have been ‘left behind’ IMHO.

    • The Z Blog says:

      How do you feel about vouchers for BMW 3-series buyers? Most people would be safer and happier driving a BMW, even a base model, than their current car. Let’s give them a voucher so they are not facing the economic barrier between them and safe transportation.

      In all seriousness, voucher schemes are doomed to failure. The private schools get hooked on the heroin of easy government money. Then the government attaches rules to the voucher, turning the private school into a public school. The people with means then move their kids to schools that don’t take vouchers.

      • Failure is a matter of degree. Regulation is a matter of degree. Granted, government money comes with strings attached. Still, overall education system performance rises as policy expands the range of education options available to parents.
        To the man with a hammer, everything looks like a nail. Devotees of HBD minimize the role of environment. Consider 1995 TIMSS 8th grade Math scores. The Singapore 5th (fifth) percentile score was higher than the US 50th (fiftieth) percentile score. If this were IQ, half the US population would qualify as severely retarded by Singapore standards. Nobody reports an Asian/white IQ gap that large. Culture matters. Curriculum matters.
        One cost of the US State-monopoly K-12 school system that appears on no balance sheet is the opportunity cost to society of the lost innovation in curriculum and methods of instruction that a competitive market in education services would generate.

        • “Devotees of HBD minimize the role of environment. Consider 1995 TIMSS 8th grade Math scores. The Singapore 5th (fifth) percentile score was higher than the US 50th (fiftieth) percentile score. If this were IQ, half the US population would qualify as severely retarded by Singapore standards. Nobody reports an Asian/white IQ gap that large. Culture matters. Curriculum matters.

          Curriculum certainly matters, but East Asians probably get a small and temporary benefit from yoking their children to desks for several years that doesn’t pay off later to the same disproportionate degree and couldn’t be replicated among other world populations, anyway. Good luck trying to get African-American kids to sit still for that many hours to do their math homework. Or even American white kids, for that matter. HBD differences aren’t just about IQ.

          When I lived in East Asia twenty years ago, Singapore was already outperforming the rest of the world in these academic tests. Yet the big story in Singapore was the lack of creativity among their smartest people. The government was already hard at work trying to solve the problem. That was twenty years ago.

          Here’s a story from five years ago: Singapore wants creativity not cramming

          HBD differences aren’t just about IQ.

      • I don’t think other people driving a BMW 3-series would improve economic outcomes for this nation in 20 years. Policy specifics aside, education does.

  2. spottedtoad says:

    I agree, although since composition has changed more for 12th graders (low achieving kids within each race are much less likely to drop out of high school now than they were forty years ago) it’s at least plausible that there’s some modest improvement to go along with the massive increase in spending.

    The funniest thing about No Child Left Behind was that when they passed the law in 2002, they actually mandated that 100% of non-Special Ed kids pass the tests by 2014. It was the Lake Wobegon law: everyone was ordered to be above average. (https://spottedtoad.wordpress.com/2016/12/31/adequate-yearly-progress/ )

  3. 17 year olds are the finished product. Best to look at the finished product rather than the earlier construction stages.

    • gcochran9 says:

      Wouldn’t hurt to see how much of the knowledge imparted in high school is retained in adults. Not all of it.

      • Mark F. says:

        I took 2 years of German and can only remember about 10 phrases.

        • Jim says:

          Some of what you learned might come back if you took up the study of German again. My father grew up in a US immigrant family where German was spoken at home. After my aunt retired she took a trip to Germany. Since she hadn’t spoken much German since she had left for college 50 – 60 years earlier she didn’t think she would be able to talk much to Germans. She thought she had forgotten most of it. But in no time at all she was speaking easily to everybody she met without any difficulty.

      • melendwyr says:

        There have been studies on that. Very, very little. That’s partly a matter of how we teach, and partly a matter of what – useless knowledge that isn’t frequently activated decays, and good riddance.

  4. Yes, given the kids that used to drop out, the flatlined 17 year old scores are actually a bit of an achievement.

    I do think we might get more of a bump if we were allowed to teach remedial high schoolers remedial skills. We could possibly push the bottom of the range up slightly. But we are legally prevented from doing so. https://educationrealist.wordpress.com/2015/08/02/education-policy-proposal-2-stop-kneecapping-high-schools/

    • gcochran9 says:

      The payoff to society wouldn’t be very high.

      • From an education outcome standpoint, maybe not. But it’s cheaper to put kids in school for 3 or 4 years than it is for them to commit crime because they have nothing to do.

        Please note I’m not saying “if kids aren’t educated, they’ll be more likely to commit crime” but rather “if kids butts aren’t in seats, they’re out committing crime.” Teachers are much cheaper than both cops and prison guards, for all the kvetching about unions. I mean, if you want to argue that a kid with a low IQ is utterly beyond hope (which I disagree with), that’s fine. But 3-4 more years in school will do some socializing and (if we aimed curriculum appropriately) give more opportunity to kids to make better choices. And if all that fails–which it will, often–then he’s not out on the streets for four years of workdays.

        I wrote about it here. https://educationrealist.wordpress.com/2013/11/21/unstructured-musings-on-choice/

        • Jim says:

          We might accomplish something useful if we tried to find out what kind of education is best for low IQ students and then gave them that kind of education. Instead we persist in futile attempts to make low IQ students “college ready”.

        • (Ed): “it’s cheaper to put kids in school for 3 or 4 years than it is for them to commit crime because they have nothing to do.”
          in Hawaii, juvenile arrests FALL in summer, when kids are NOT In school. Juvenile hospitalizations for human-induced trauma FALL in summer, when kids are NOT in school. Juvenile arrests for drug possession and drug promotion FALL in summer, when kids are NOT in school. Adult arrests for possession rise in summer. Adult arrests for promotion fall. The juvenile drug market evaporates and the supply shifts to adults.
          Schools don’t prevent crime; they cause it.

          • Actually, it’s true that schools have criminalized a lot of behavior. But mostly fine-based stuff. The fact that it’s easier to catch drug users during the school year doesn’t mean schools cause crime, and this whole line of argument has nothing to do with whether or not there’d be more crime if unmotivated kids could drop out.

        • Still, having criminals in class is pretty expensive to actual students.

          I know it isn’t politically feasible, but why not just make prisons cheaper? I’m sure Arpaio style tent prisons compare better to schools than most US prisons do.

        • Democritus Junior says:

          You’re arguing for the social advantages of babysitting teenagers? Your whole pitch is that you’re cheaper than a parole officer and an ankle bracelet? (Actually, I’m not sure that the numbers on in your favor there.) I assume that you only think that the high risk populations need to be babysat in this way, or is everybody else just collateral damage?

          In fact, most of the behavior effect from school does not come from teachers, it comes from peers. School has a huge anti-socialization effect. If you want to do something positive for the world, get young adults out in the working world where they can socialize with adults and integrate into their community rather than the false and poisonous community of school.

          • No, arguing for the financial advantage. And my argument is that if you assume all kids incapable of being educated past fifth grade are hopeless, then leaving them on the streets without jobs will lead them to crime. Crime, violence, etc is a lot more than a parole officer and an ankle bracelet. There’s the additional cost of policing, the impact 0on the victims, the increased prison requirement, etc.

            “School has a huge anti-socialization effect.”

            Wrong.

            “If you want to do something positive for the world, get young adults out in the working world where they can socialize with adults and integrate into their community rather than the false and poisonous community of school.”

            hahahaha. Yeah, what jobs would those be, exactly? We’re talking about kids who are, by your definition, unteachable. They won’t be working and getting socialized. If they were capable of getting jobs, they’d be capable of minimal performance at school and thus better off for school (stats show). There’s no jobs for the kids we’re talking about. When these jobs exist, I agree that they should be allowed to work.

            • engleberg says:

              <<Median charter schools and private school voucher programs don’t meaningfully outperform traditional public schools. High-performing charters are generally not remotely scalable. Charter schools writ large are rife with mechanisms that screen out the hardest-to-educate students, even when they advertise as open to all. Schools like the Harlem Children’s Zone schools are notorious for sky-high student attrition rates. Systems like the Success Academies rely on a constant churn of teachers coming into and out of the system in a regular and predictable burnout cycle; this in turn requires a steady influx of educated labor willing to work in poor conditions, a situation that cannot possibly be ported from New York City to, say, the Mississippi Delta. According to charter advocates hundreds of thousands of teachers across the country would have to be fired, but we already face a teacher shortage and there is no bullpen of teachers ready to step in. Attacks on tenure make the job of public teacher less attractive; claims that this will be mitigated with increased teacher salaries depend on utterly fanciful notions that we’ll dramatically raise property taxes across the country. Charter and private schools have direct financial incentive to commit fraud to raise standards. I could go on.>>

              Freddie de Boer, not a charter fan. It looks to me like charter schools blow some smoke about diversity, then Expel Thugs. Haven’t asked Smoke.

    • savantissimo says:

      More of a bump if they could be convinced to drop out, but that would be gaming the numbers. The real cost of NCLB is not allowing the top students to progress at the rate that they could. The tests don’t have enough “top” to show what’s going on in the top 2%, let alone the top 0.2% of students who matter more for advancing civilization than the bottom 20%.

    • pyrrhus says:

      The Europeans used to have a system, Swiss may still have it, where kids who weren’t students were pushed out into the trades at about age 15, which allowed them to get a head start on making a decent living. What we are doing now is warehousing kids who have no interest/aptitude in academic subjects through high school, and even at the junior college level. Princeton now offers remedial courses to its AA students, such as the “wise Latina”, who says she was encouraged to read childrens books to improve her reading.

  5. savantissimo says:

    Here are the graphs: https://nces.ed.gov/nationsreportcard/pubs/main2012/2013456.aspx

    I avoid reading education research, especially from education advocacy, the idiocy and lies make me feel tired and angry. A friend from high school went on to become one of Teach for America’s C-level executives, in charge of spinning stats. I can’t bring myself to talk to him, the BS he spouts is so thick. (TFA is now a prestigious thing for Ivy League grads to put on their resumes, wasting two years attempting to teach the stupidest brown kids they can find. What a waste.The teacher training is graduate-level PC indoctrination. David Gergen, Joel Klein, Ken Mehlman, Larry Summers and Meg Whitman are the sort of people they put on their board.)

    One interesting stat from an outside study of TFA: grading teachers by their students’ improvement on tests, a +1 s.d. teacher has students that do +0.2 s.d. better than their prior scores. It would theoretically take a +5s.d. teacher (1 in 1,000,000+) to bring the -1 s.d. average Black up to the national average. (Except that the improvement is probably really just statistical noise.)

    • RCB says:

      Regarding that last stat: I’m guessing that “teacher quality” was defined by classroom improvement in whatever study this was. I.e., a teacher was called a “+2 sd teacher” if their class room’s improvement was in the top 2 sd of the class room improvement distribution. Which would be pretty meaningless by itself. Do you know if that’s true?

      • spottedtoad says:

        Yes, that’s correct- they define “a good teacher” as one whose students improved in scores more than expected in a given year. I wrote a little about why I think this kind of value-added modeling is genetically confounded here: https://spottedtoad.wordpress.com/2016/02/03/value-added-modeling-and-behavioral-genetics/

        • RCB says:

          Even knowing nothing about genetics, it’s meaningless. Randomly assign kids to classrooms, and give them common inanimate household objects for teachers. By chance, some classes will improve more than others. We’ll conclude that Spatula is a better teacher than Mousetrap.

          Statistically, if students were divided randomly and there were no difference in teacher quality, then the variance in classroom improvement would be equal to the variance in individual student improvement divided by the classroom size. So that provides your null hypothesis. I’m sure some researchers have addressed this…

          • spottedtoad says:

            Yeah, I think pure randomness would produce a classroom-level component of variation that shrinks with the square-root of class size, but in practice the classroom-level component of variation is larger, because kids aren’t perfectly randomly assigned even within schools (and because classrooms differ in experience, quite apart from teacher efficacy), and because individual kids converge to different set-points over time, that can’t be fully captured by baseline achievement.

      • savantissimo says:

        Yes, teachers’ scores are ranked against the distribution of classroom test-score improvements. No reason to think that such gains are anything but random since there is no data on whether teachers’ score on that metric have any consistency year-to-year. Also the test improvement numbers for a class can easily be raised if a few students under-performed on the baseline test compared to their ability, or lowered for bad scores on the second test, so by sheer chance some teachers will look good and others bad without any real difference in their teaching.

        Still, the point that teaching can’t be shown to have much effect on student performance is only reinforced. Student quality, primarily intelligence, but also work ethic determines student performance, and student quality is mostly genetic and would be more so if students were allowed to select their own environments, but the variation caused by environment can’t sustainably improve over what is possible genetically, it can only cause variation by limiting students to less than their genetic capacity, but fortunately generally only temporarily.

  6. uncommonman says:

    Look, you unimaginative brutes. There are a LOT of things we can do to improve general education (higher ed needs no help). I haven’t finished my coffee, and I can think of ten just of the top of my head: Discipline, parental involvement, discipline, discipline, discipline, discipline, discipline, discipline, discipline, and discipline.

  7. Pingback: Liberty and Social Norms Re-examined | POLITICS & PROSPERITY

  8. Jim says:

    The ability of governments to persist in totally irrational policies is truly amazing. Greg mentioned Soviet steel production. Today another Greek bailout seems to be in the works. It would have been far more rational to devote the hundreds of billions of euros spent on “bailing out” Greece to a project to develop a perpetual motion machine. The latter undertaking might have produced some useful “spinoffs”.

    • Greying Wanderer says:

      “Today another Greek bailout seems to be in the works”

      They’re not trying to fix Greece. They are keeping it on life support so they can loot it more.

      By the time they’ve bled the corpse dry every banker will have their own privatized Greek island.

  9. Mark F. says:

    Question: What would be the impact of the total abolition of free public education? I’m not any libertarian ideologue, but am just wondering what you folks think would happen.

    • I am not a libertarian- they should all be shot- but I can offer an opinion on what would happen.

      Assuming that the tax savings were passed on, and no further restrictions were imposed on freedom of association, things would improve for the majority of decent parents. They would be better able to afford private, but sensibly priced ‘academies’, as flourished in the south after mandatory integration was imposed on students and parents. Property price distortions would be reduced, as physical separation became less important to separate your children from bad students.

      Status-signalling billionaires would have to shift to ‘paying for bad students/bad schools’ rather than ‘paying to chuck useless ideas at bad students/bad schools’.

      Of course, if we assume that this could actually happen in reality, it could only happen in a reality where the bill that abolished public schooling would mandate that parents who did not pay to school their children, or fill out the Gates Foundation Standard Grant Form, could be jailed and sterilized.

      Basically, it would never happen, but if it did happen, it would be great. It would just require such a sea change in politics that you can’t look at such ideas out of context.

      • No,things probably wouldn’t improve for “decent parents”. Private schools free-ride off of the higher paying public schools–they get most of their staff from people who never bothered to get a teaching credential or already taught in public schools and are willing to accept low pay. Public school teachers get paid much more. If all parents were going to private schools,teacher supply would be a real problem. That will drive the cost up, making private schools a hell of a lot more expensive.

  10. Agent J says:

    My guess is, what you’re seeing is probably around half due to the distribution of IQ and half due to issues involving socioeconomics and teaching. I doubt you could do much more than maybe 20-25% better, no matter what changes you make in schools or how much money you dump down the public education drain. Issues of SES and IQ will put an upper limit on any improvement.

  11. Anonymous says:

    FYI, you see to have posted this wrong. Look at the black bar up top.

  12. dux.ie says:

    Some children leap ahead.

    I tried to find PerK the density of the semifinalist National Merit Scholars NMS in CA cities with respect to some measure of population less than 25 yo Pop24 . (Student numbers are hard to find and the participation rates are unknown). NmsNdx is the ratio of PerK over that of the state average.

    NmsNdx PerK Est% NMS Pop24 State City
    57.02 14.88 35.71 5 336 CA Ross
    49.08 12.81 30.74 29 2264 CA Atherton
    16.89 4.41 10.58 24 5444 CA Westlake Village
    16.82 4.39 10.53 89 20277 CA Cupertino

    9.03 2.36 5.66 74 31383 CA Palo Alto
    6.72 1.76 4.21 37 21082 CA Mountain View
    5.35 1.4 3.35 27 19321 CA Arcadia
    2.08 0.54 1.3 6 11039 CA Beverly Hills
    1.0 0.26 0.63 2053 7866048 CA State average
    0.8 0.21 0.5 9 43304 CA Berkeley
    0.67 0.18 0.42 32 181738 CA San Francisco
    0.09 0.02 0.06 20 829589 CA Los Angeles

    • dux.ie says:

      NmsNdx is positively correlated to the social-economic status percentile SESpctl, median income (in ‘000) IncK and the fraction of degree holders. The three factors are also correlated among themselves, implying there are common latent factor(s) among them. For example the respective values for the top four cities are,

      SESpctl 99.94 99.99 95.74 98.72
      IncK 228.33 250.0 131.67 154.13
      FracDeg 0.87 0.84 0.6 0.76

      Their effects on NmsNdx are non-linear, thus the effects on log(NmsNdx)
      are considered.

      At first glance SESpctl could be the dominant factor, as those SJWs asserted. However, regression of log(NmsNdx) with the three variables shows

      Coefficients:
      Estimate Std. Error t value Pr( gt |t|)
      (Intercept) -2.992113 0.234928 -12.736 lt 2e-16 ***
      SESpctl 0.001629 0.006291 0.259 0.796
      FracDeg 5.248951 1.036215 5.066 8.97e-07 ***

      IncK 0.003983 0.003666 1.087 0.278

      Signif. codes: 0 ‘’ 0.001 ‘’ 0.01 ‘’ 0.05 ‘.’ 0.1 ‘ ’ 1

      n=212; Rsq=0.5758; p lt 2.2e-16

      With FracDeg present, SESpctl and IncK are simply not significant. FracDeg is the representative cognitive ability of the parents while the other two are representatives of the environment. Thus genetic is the dominant factor for determining the NMS semifinalists, the reverse of what the SJWs asserted.

      Furthermore, these semifinalists are the top 1% in terms of the cognitive ability of their peers, the furtherest from the population mean. If regression to the population mean is dominant, they should be the group experiencing the greatest pressure, yet they are able to persist as a cluster density as large as 57 times the average. It is more like the regression to the mean with respect to their own sub-population clusters rather than to the overall population.

      Murray had now asserted that due to assortive mating, the IQ distribution even for the same ethnic group should be bimodal (multi-modal?). The above statistics shows that those semifinalists most probably are the results of of people with close SESpctl, IncK and FracDeg staying in close geographical locations, and stayed seperate from the overall population.

    • savantissimo says:

      Ross and Atherton both have very selective private schools that are large compared to the supposed population (larger than the supposed under 24 population in the case of Ross) and draw from surrounding areas outside the town. Crystal Springs Upland School in Atherton claims 2004 SATs of 1460-1600 with 243 in the upper school, so should have close to 60 semi-finalists by itself. Neither town has public high schools of its own, so far as I can tell. Those two data points are not comparable to the others.

    • dux.ie says:

      There is no surprise where is the top cluster in New Mexico,

      NmsNdx PerK Est% NMS Pop24 SESpctl FracDeg IncK State City
      11.4 2.45 5.88 13 5308 95.69 0.64 124.05 NM Los Alamos
      5.58 1.2 2.88 1 834 88.31 0.55 93.16 NM Sandia Park
      1.81 0.39 0.93 1 2576 38.01 0.29 43.89 NM Taos
      1.16 0.25 0.6 1 4015 20.91 0.22 37.59 NM Socorro
      1.13 0.24 0.58 53 219062 50.66 0.32 60.76 NM Albuquerque
      1.0 0.21 0.52 85 395720 50.25 0.32 60.65 NM state average <–
      0.81 0.17 0.42 6 34486 64.95 0.41 66.61 NM Santa Fe

  13. DataExplorer says:

    What do you think about early intervention for children with Autism? In home ABA therapy is all the rage these days. Does it have any lasting impact?

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s