O'Neil Perspective2011-07-11T19:36:12-04:00lifetype-1.2.11_r7114http://www.oneilresearch.com/lifetype/lt/rss.php?blogId=2&profile=atomCopyright (c) mikeoneiltag:www.oneilresearch.com,2011-07-11:140Joplin . . . and the Role of Government2011-07-11T19:36:12-04:00
A few days ago, I
watched the Joplin Memorial Ceremony attended by President Obama, the
governor of Missouri and a host of others. Something has stuck in my
craw since viewing ...mikeoneil
General
A few days ago, I
watched the Joplin Memorial Ceremony attended by President Obama, the
governor of Missouri and a host of others. Something has stuck in my
craw since viewing that ceremony. . . .
Click here to see the remainder of the article in the Huffington Post .
Or click here to listen to a recent comprehensive radio interview on politics and polling.
tag:www.oneilresearch.com,2011-04-18:139Tax Day Observations: An Immodest Proposal2011-04-18T14:32:55-04:00
Tax Day is today. Unless you are an anarchist, you recognize that taxes are a necessary evil. Most discussions about reforming the tax code, however, conflate two ...mikeoneil
General
Tax Day is today. Unless you are an anarchist, you recognize that taxes are a necessary evil. Most discussions about reforming the tax code, however, conflate two entirely separate issues:
1) the amount of taxes we pay, and
2) what it is that we tax.
Let's set aside for today, the question of how much we tax. There is great illogic in WHAT we choose to tax . . .
Click here to see the remainder of the argument in today's Huffington Post .
tag:www.oneilresearch.com,2011-04-04:138Why Most Internet Surveys are Useless2011-04-04T19:42:39-04:00 When I was about five years old, my mother told me to that it was too dangerous for me to play in the street. "But mom, ALL of the other kids' mothers let them play in the street! ...mikeoneil
General
When I was about five years old, my mother told me to that it was too dangerous for me to play in the street. "But mom, ALL of the other kids' mothers let them play in the street! It MUST be safe!" Mom was, of course, right--the alleged behaviors of all of the other moms notwithstanding.
But replace "moms" with "major corporations" and we get deference to the dumbest unsupported practices imaginable.
Independent Proof
A recent report by the US Census confirms what I have been advising clients for years: you simply cannot trust the vast majority of internet polls, since they are based on opt-in volunteer samples that are not representative of the broader population.
It has been two generations since the "Dewey beats Truman" survey debacle in 1948 demonstrated the frailty of nonprobability methods. (The University of Michigan was the only institution employing probability sampling that year; it was the only survey organization to get that one right). Since then, it has been demonstrated over and over that the only way to ensure that the results of any survey will be projectable to a population is with a probability sample . (Ouch! Technical Term. Sorry, but it is necessary to understand the point). And the principles have been validated in tens of thousands of surveys for over 60 years.
Probability samples work for a reason : they are the only sampling methods firmly based on the laws of mathematics. These principles also are the basis of the ubiquitous "plus-or-minus" calculations. Without a probability sample you cannot even compute any margin of error. So says the leading professional organization in the field (www.aapor.org). Of course, the purveyors of such methods "compute" these anyway; but they are garbage numbers based on fantasy.
You cannot get a probability sample of the general population in an internet survey . And not just because some people don't have email addresses, but because there is no way to compute probabilities of selection for the remainder who do. (OK, that's technical again, but it is a critical distinction). Internet sampling methods, therefore, have NO basis in mathematics. In other words, NO reason to believe that they are worth anything.
But perhaps, all this doesn't matter? Maybe these internet polls work anyway? After all there are so many of them! And, oh so many moms - oops, I mean large corporations - use them that they must be OK, right?
Usually you have to have a defensible basis for believing that something may work in order to justify the effort to test it. While internet surveys have no scientific basis, the sheer sums being spent on them have made this an interesting question. And so, the largest survey research organization in the world, the U.S. Census, recently commissioned just such a study. The university researchers who conducted the study are respected professionals without an axe to grind. The gist of their results: internet surveys simply don't pass muster. A surprise to virtually no one with a modicum of statistical training, but it provided definitive empirical evidence for that which was already obvious to the statistically literate.
(For the numbers geeks reading this, I provide a link to the entire report . For those who want a briefer summary, I provide a link to an excellent and brief journalistic explanation of the results).
Then Why Are There So Many Internet Surveys?
Why, then, the proliferation of such surveys? There are two compelling reasons why corporations have come to use internet surveys:
1. they are quick, and
2. they are cheap.
Since statistical literacy is seldom taught in business school (really, it isn't), this set the stage for an entire industry to grow up based on assuring corporations that their quick and cheap methods can be relied on. And, because quick and cheap are both compelling attributes (who doesn't like quick and cheap?), many of the largest corporations in the world want to believe that such surveys are OK. "I want the answer NOW and I don't want to pay much for it!" demands the "decisive" corporate executive. (Kind of the same corporate bravado that lets a major corporation build a nuclear plant on an earthquake fault line: it seems smart - at least until something happens).
And an entire multibillion dollar industry has developed to meet the demands of quick and cheap. Just don't look too carefully at the foundations of what quick and cheap really gets you. And these are not just private corporate surveys, they have worked their way into the publicly available surveys on politics and public issues--and they are equally invalid for either purpose. (Fair acknowledgment: the three major networks have started to do a pretty good of screening this garbage out. But the local press' and the blogosphere's record is poor; they simply don't have the trained staff to separate the good from the bad).
But aren't the alternatives to internet surveys imperfect? Don't some people refuse to be interviewed in any method? And don't people sometimes lie to pollsters? Yup, on both counts. But there are usually pretty good ways to handle both of these problems. Fixes that are based on decades of research, not a wish and a prayer. The methods are technical and involved (and therefore boring to all but survey geeks). But they do exist. Unfortunately, the fixes are neither quick nor cheap, so most companies don't really want to hear about them. There is a Grand Canyon sized gap between that which is baseless and that which is merely imperfect.
Note: the problem is not with the use of the internet per se , but with the fact that most internet surveys have no choice but to use samples of those who volunteer to be interviewed, usually for a modest fee or prize. These self-selected volunteers are the only choice available to internet researchers. There is a single exception: surveys of membership organizations when virtually all of the members have email addresses available to the organization and the membership is sufficiently committed to the organization to be willing to participate in a survey. We have done such surveys frequently. But these are a tiny slice of the internet surveys that are reported in the press.
Save this article. The first multibillion dollar corporate mistake based on the use of internet surveys is coming. I guarantee it. Indeed, it has probably already happened. But the affected company may not know about it yet-or may not be anxious to admit such a mistake. When the mistake becomes obvious, these same "decisive" corporate executives will demand to know "Why?" But, we already know the answer.
tag:www.oneilresearch.com,2011-02-10:137Senator Kyl Retires2011-02-10T20:46:32-05:00 Early speculation on who's likely to be in the race to replace Senator Kyl: click here for Fox10 video. Michael J. O'Neil, PhD mikeoneil
General
Early speculation on who's likely to be in the race to replace Senator Kyl: click here for Fox10 video. Michael J. O'Neil, PhD
tag:www.oneilresearch.com,2011-01-16:135Egypt: Now the Hard Part2011-01-16T20:34:14-05:00 Many thousands of Egyptians, fed up with 30 years of oppressive rule, successfully called for Mubarak's resignation. Their rhetoric is all about Democracy. If their focus remains only on Mubarak, ...mikeoneil
General
Many thousands of Egyptians, fed up with 30 years of oppressive rule, successfully called for Mubarak's resignation. Their rhetoric is all about Democracy. If their focus remains only on Mubarak, however, they're unlikely to end up with anything resembling real democracy.. . . unless other institutions are in established and new norms of civic behavior are operative, the version of democracy that Egypt may end up with may resemble that which we have seen in numerous countries elsewhere in Africa: "One Man, One Vote, One Time." See the rest at the Huffington Post. Click here to view: http://www.huffingtonpost.com/michael-j-o/egypt-now-the-hard-part_b_822583.html Or, for some observations on Senator Jon Kyl's seat and possible candidates for it, click here . Michael J. O'Neil, PhD
tag:www.oneilresearch.com,2010-12-13:134Free Lunch!2010-12-13T17:17:37-05:00
I'd love a tax break! Pay less! And that's just what the President
and the Republicans in Congress just worked out. And they are going to
pay for it with. . . ...mikeoneil
General
I'd love a tax break! Pay less! And that's just what the President
and the Republicans in Congress just worked out. And they are going to
pay for it with. . . No, wait! The deal was sweetened with INCREASED
for unemployment benefits. . . and?
I think we now know the formula for split party government: LESS taxes and MORE spending! $900 Billion dollars worth.
....see the remainder of the posting, on today's Huffington Post: http://www.huffingtonpost.com/michael-j-o/id-love-a-tax-break_b_794996.html
Michael J. O'Neil, PhD
tag:www.oneilresearch.com,2010-12-06:133Stuff and Nonsense About Taxes2010-12-06T22:43:13-05:00 As the decade of the 1970s was ending a one-time Saturday Night Live comic deadpanned an ongoing routine that he introduced as follows: “Well, the ‘me’ decade is ...mikeoneil
General
As the decade of the 1970s was ending a one-time Saturday Night Live comic deadpanned an ongoing routine that he introduced as follows: “Well, the ‘me’ decade is almost over, and good riddance, and far as I'm concerned. The 70's were simply 10 years of people thinking of nothing but themselves. No wonder we were unable to get together and solve any of the many serious problems facing our nation. Oh sure, some people did do some positive things in the 70's - like jogging - but always for the wrong reasons, for their own selfish, personal benefit. Well, I believe the 80's are gonna have to be different. I think that people are going to stop thinking about themselves, and start thinking about me, Al Franken. That's right. I believe we're entering what I like to call the Al Franken Decade." ....see the remainder of the posting, on today's Huffington Post: http://www.huffingtonpost.com/michael-j-o/stuff-and-nonsense-about-_b_792693.html Michael J. O'Neil, PhD
tag:www.oneilresearch.com,2010-11-03:132What's Next: Two Scenarios2010-11-03T12:56:23-04:00
OK, the Republicans won big in a "wave" election. What happens next? I have two scenarios
The Optimistic View
Both parties recognize that they have to compromise in ...mikeoneil
General
OK, the Republicans won big in a "wave" election. What happens next? I have two scenarios
The Optimistic View
Both parties recognize that they have to compromise in some matters
of policy. There are two obvious ones from which each side could gain
something . . .
Click here to read the rest in today's Huffington Post .
Michael J. O'Neil, PhD
tag:www.oneilresearch.com,2010-11-02:131Everything You Need to Know About Tonight's Election2010-11-02T14:20:54-04:00
Education . The piece I sent you last week was published in the Huffington Post . Created a minor firestorm of comments. Evidently pricked some sensibilities among that readership. ...mikeoneil
General
Education . The piece I sent you last week was published in the Huffington Post . Created a minor firestorm of comments. Evidently pricked some sensibilities among that readership. Regardless of your reaction, I'd love it if you would add your comments to that lively discussion. Click here to do so.
Election Broadcasts . I'll be on Fox 10 (Phoenix) election night and Horizon (PBS, 8) on Wednesday. But, if you want to keep me honest you can click here to see some of my earlier prognostications on the election-starting from things started to heat up earlier in the year.
Election Night Forecast. We're
really more in the analysis than the prediction business, but the world
doesn't accept that, so I will play that game and tell you exactly what
is going to happen so you can go to bed early tonight. Print this out
and see if I have nailed it.
House. Republicans pick up 55 seats-plus or minus 20. Since
they need 39 to gain control of the House, that means that the outer
reach of Democratic hopes are that they maintain a slim majority
(unlikely, but possible). Chinese Proverb: Be careful what you wish for. With
the Republicans in control of the House, the next election will be
about which party people prefer, not a referendum on the Economy or
Obama's stewardship of it. The Republicans have made this election about the latter, and that's why they are winning -- and big . If the election were about which party people like more right now, the Democrats would actually win, but narrowly. The Republican Party is actually less popular than either the President or the Democratic Congress, but that will not matter tonight.
U. S. Senate. I'll get specific: There are only 10 seats in real contention. All are held by Democrats. The
Dems will lose three with certainty (North Dakota, Arkansas and
Indiana), two with near certainty (Wisconsin and Pennsylvania), and
will probably lose Nevada as well. (Though note the slight changed in adjectives, that's deliberate).
Now it gets fun. I'd rate the Republicans as slight favorites to take Colorado and Illinois. If they fail to carry both, Senate takeover hopes are toast. Give them these and they are within one seat of an evenly divided Senate. (While Joe Biden breaks the tie, don't count on that holding. With a 500-50 Senate either Joe Lieberman or Ben Nelson could jump ship. That scenario would be both fun to watch and unseemly. The Democrats really need 51, not 50).
So we have the end result: the Republicans end up with 49 Senators. (Keep an eye on who might get sick). My guess is that's where they stay. They
have a reasonable shot in West Virginia and Washington, but not quite
as good as the Democrats have of holding Colorado or Illinois).
The Democrats chance of winning any currently Republican held Senate seat: Near Zero. This is a wave election, like 2006 or 2008. What
will make the numbers somewhat bigger this time is that many of the
seats the Republicans will pick up are low hanging fruit: the seats
they lost in the last four years.
Wild Card . Really screwy things are going on in Alaska. May
well provide post-election night entertainment (think Al Franken's
election which took weeks to count all the ballots) after which some
Republican probably ends up with the seat. But a lot of people can see
all kinds of things from their back yards there, so who knows?
The 2012 Presidential Election will start no earlier than Wednesday morning. Maybe Tuesday night. Something to chew on: a Republican takeover of the House will help Obama's prospects for re-election. He will no longer have sole ownership of the economy, which is likely to be improving some but still sputtering in two years.
For the Locals .
The only interesting races will be the Congressional Seats . Five of the eight (more than in Arizona history) have at least an interesting story. Republicans
sweep everything else in sight. (If there is an exception, it would be
Felecia Rotellini: will the voters want a Democratic Attorney to keep
the Republicans who will control everything else honest? Maybe.)
Here's the congressional story. Three seats are safe : Ed Pastor (D, CD2), Trent Franks (R, CD4) Jeff Flake (R, CD6). Don't even bother counting the votes: they have already won.
Three seats currently held by Democrats that are all traditionally Republican seats . All were swept by Democrats in one or both of the last two Democratic sweep elections: Anne Kirkpatrick (CD1), Harry Mitchell (CD5), and Gabrielle Giffords (CD8). All three Democrats are in jeopardy. All the races are close. My
guess is that Giffords hangs on, but that Kirkpatrick does not. The
Mitchell race is the closest of the three, but most national
prognosticators give his challenger Schweikert the edge. Mitchell
is a beloved figure in the District, and until this election, not seen
as a particularly partisan one. But given the Republican tidal wave,
will this be enough? They'll actually have to count the votes on that
one.
Fantasy Wins. Each party has a chance to win in a district where they would normally have no chance:
CD3 :
Ben Quayle has been an extraordinary weak and flawed candidate,
providing an opening in a district that is as safe a Republican
district as they come. If elected, however, John Hulburd would probably
be among the most conservative Democrats in the House. And,
I'd bet he'd serve a single term: he could be beaten by a generic
Republican in two years; this is way too Republican a district for any
Democrat to hold.
CD7 :
Raul Grijalva supported a very unpopular boycott of Arizona. This
district is the flipside of CD3, if people are peeved enough to vote
him out for this reason (don't bet on it), his successor will serve two
years and be beaten by any Democrat in 2012.
Neither of these upsets would have anything whatsoever to do with national trends. But
the three competitive races, Kirkpatrick, Giffords, and Mitchell could
be national bellwethers. If the Democrats hold all three, the
Republican wave will be on the small side. If the Republicans carry all three, that wave is BIG. My guess: split decision.
The only Blue State victory over a Red State this week was when the San Francisco Giants beat Texas yesterday. The only suspense will be how big the puddle of blood is on the floor tonight.
But
it could be short-lived. We have national political ADD: the Dems held
the House for 40 years, and then the Reps for 12, the Dems then got it
back for 4. At that rate, the trend would suggest that the Dems will be back in 2.
Print this out and see if I nailed it. Accountability.
Micheal J. O'Neil, PhD
tag:www.oneilresearch.com,2010-10-29:130Education Reform and the 2010 Election: Two Lousy Choices2010-10-29T20:01:54-04:00
I have had my 2010 election ballot for almost a month now. I found most of the choices to be fairly easy. I
have agonized, however, about one particular choice. I'd like to ...mikeoneil
General
I have had my 2010 election ballot for almost a month now. I found most of the choices to be fairly easy. I
have agonized, however, about one particular choice. I'd like to share
my thinking with you, not to urge you to vote one way or another, but
to explore the relevant issues behind my dilemma.
You have probably received endless communications at this point advocating various candidates and positions. I
want to share with you not so much my advocacy of a particular
candidate, as the deep ambivalence that I feel toward the choices that
we voters have to make for one office, the State Superintended of
Public Instruction. (The office is in Arizona, but all of the points are relevant throughout the country).
I have read the election materials of each candidate, read their election pamphlet essays, and watched them debate. While
they have substantially different positions, I am absolutely convinced
that neither will do anything whatever to alter the pathetic status quo in American public education.
While
this is an Arizona election, the candidates are almost caricatures for
the two major strains of thinking about what is wrong with American
public education:
The
Republican candidate, a state legislator, was an active participant in
substantially reducing the level of funding for public education in the
state. From this I infer that he thinks we can improve our schools with less, rather than more, resources. His position papers include all of the usual Republican canards, "high standards" and references to "English-only education." The magic pill that he and many of his ilk offer is that of charter schools. If
we simply encourage private entities to create their own schools, these
will compete with public schools and will solve all of our educational
problems-with no increase in resources.
The Democratic candidate is also typical of the Democratic view of most of these issues. As
such, she decries any cuts in educational spending, is silent or
lukewarm on charter schools and generally thinks that the answer to our
educational mess is just supporting the teachers that we have-with
increases in funding. But as head of the teacher union,
she held a position whose central tenant seemed to be to protect the
job security of the most incompetent teacher who ever walked into the
classroom without any regard for the impact on students.
I watched both of these candidates in a public debate. Superficially, each seemed articulate, intelligent and reasonable. If
I didn't feel I was able to read between the lines of their positions,
I might have felt that either could improve public education. When I
looked beneath the surface, however, what I saw was a pair of
candidates, each completely beholden to the interests of their
respective constituencies. The more I thought about
it, the more I was absolutely convinced that there is not the slightest
chance that either would contribute in any meaningful way to improving
public education.
Much
has been made of the fact that American students are performing at a
level near the bottom among its peers, the other industrialized nations. In looking at these discrepancies and the differences, one fact stands in stark contrast. The
most successful counties pull their teachers from among the top third
of college graduates, while most of our teachers are pulled from the
bottom third. And both these countries and ours pay accordingly: they pay a lot more than we do. And value their teachers accordingly.
This
stark fact suggests to me an obvious solution, one I am certain each of
these candidates would reject out-of-hand (but for different reasons). Let's announce that in the state of Arizona, effective four years from next September we will double the salaries of all public school teachers. (I am sure the Democratic candidate would love this.) Oh,
but on that date, teacher tenure would end; all current teachers would
lose their jobs, but be free to compete to be re-hired for their now
much higher-paying positions. (Now the Republicans cheer). But, they
would have to compete for their old jobs with a generation of recent
graduates who would go through school knowing that Arizona values
educators and is willing to pay them accordingly. These
would include students who would otherwise consider such professions as
law, medicine, biosciences or engineering; the most challenging
subjects that currently attract our best and brightest students. And
current teachers would also have to compete with the best teachers from
all over America, and perhaps the world.
And
in doing this hiring, I would favor persons who had excelled in the
subject matter which they teach, rather than education. The dirty
little secret is that education is the easiest subject in which to
major in every college and university of which I am aware.
And
to keep these now highly-paying jobs, these newly hired teachers would
have to continue to deliver. Exactly like the doctors, lawyers,
engineers and scientists the best of whom are currently paid so much
more. (The untold side of the lament that we pay teachers so poorly is
that even the worst of them is guaranteed a paycheck for life: a
paycheck equal to that received by the very best of their fellow
teachers. All the highly paid professions offer NO guaranteed salary. And the worst of the lot in these esteemed professions can actually earn less than the wages we guarantee for life to all of our teachers, regardless of their performance).
Is this the be-all and end-all of the educational reform we need? Of course not. I'll say more about the rest of the story in future missives. But I do offer this as food for thought.
In
the meantime, there is actually some hope for the employment prospects
for Arizona schoolchildren-for jobs for which they will fully qualify
under our current educational regime. If Russell Pearce gets his way, there will be a lot of vacancies for positions as maids and gardeners in the state.
Oh,
a few minutes before writing this, I finally marked my ballot for
Superintendant of Public Instruction: I cast a write-in vote for Michelle Rhee .
Michael J. O'Neil, PhD