BlogaBarbara

Santa Barbara Politics, Media & Culture

Thursday, November 15, 2007

Community Post: Suicide Barriers Alternative on Cold Springs Bridge

Unfortunately, I don't have a way of linking to the actual, alternative proposal, if anyone has one -- please pass it along.

--------Friends of the Bridge Community Post/Press Release-----------

Opponents of a CalTrans proposal to install barriers on the Cold Spring
Bridge today formally submitted to that agency a detailed alternative
project proposal that they claim provides superior suicide prevention and
life-saving measures at a small fraction of the $1,000,000 allocated for
barriers and which would not require potentially-defacing physical
alterations of the bridge. By law CalTrans will be required to incorporate
the alternative into its project review and evaluation processes and to
make a carefully considered determination about which approach to pursue.

The new proposal is based in part on the study recently released by UCSB
Professor Garrett Glasgow revealing the lack of any evidence that bridge
barriers do anything other than diverting the problem elsewhere. The
Glasgow report concludes that "there is no evidence that a suicide
prevention barrier on the Cold Spring Bridge would save lives."
Accordingly, the new proposal asserts that "the barriers proposal amounts
to nothing more than a misguided and costly suicide diversion effort that
bears almost no legitimate relation to any traffic safety problem within
the scope of CalTrans' mission and primary functions."

The new proposal discloses that the $1,000,000 allocated for barriers was
taken from funds budgeted for "Collision Reduction-Safety Improvements."

In formulating the new proposal, its authors from the newly-formed citizens
group "Friends of the Bridge" incorporated a number of the features of a
plan adopted earlier this year by the New York State Bridge Authority to
address the problem of suicidal behavior on the 5 bridges under its
jurisdiction. The New York plan rejected the idea of installing barriers,
choosing "instead [to] construct a 'human barrier' that will outperform any
physical barrier and save more lives" and concluding that "preventing
suicides on [the] bridges will most likely occur if we recognize the
situation for what it is: a mental health problem that won't be solved by a
technical 'quick fix' in the form of a 'curtain of steel'...."

Elements of the "human barrier" approach now proposed for the Cold Spring
Bridge include setting clear policies and training methods to assure the
safety of law enforcement officer and others when encountering persons
demonstrating suicidal behavior on the bridge, installing call boxes
connecting to a specialized Lifeline suicide prevention counseling service,
and installation of surveillance cameras to provide continuous monitoring
of activities on the bridge.

The proposal points out that its features could be fully implemented within
a very short period, whereas CalTrans has estimated that the barriers
proposal faces many months of review and design efforts. According to Marc
McGinnes, one of the authors of the new proposal, "The jig is up, we now
know for sure that barriers are boondoggle, and the time has come to
CalTrans to change its course."

Labels: ,

59 Comments:

Anonymous Don Jose de la Guerra y Noreiga said...

There it is folks. Leave the bridge alone. There's good planning for you.

11/15/2007 9:37 PM  
Anonymous Anonymous said...

A barrier a boondoggle? What has McGinnes been smoking or drinking. The 10 touchy-feely half-measures proposed by this group is the real boondoggle.

The gameplan of this group is to complexify, delay, inflate the cost, and ultimately kill a project that is proven (by the Seiden study) to save lives.

The Glasgow study they refer to would get a D in any serious upper-division statistical methods course.

They are very clear: the world is better without the people who commit suicide off the bridge. They don't value human life. They are the analogues of Bush's torturers on the left wing.

Let Caltrans solve this issue and save some human lives that McGinnes and is buddies believe are worthless.

11/16/2007 4:39 AM  
Anonymous Observer said...

Hmmmm... This presents a difficult philosophical divide.

Do I want "cradle-to-grave" protections of every description to save me from myself or do I prefer that the ACLU argues here for my right to kill myself!!?

Oh, choices, choices. Please someone, anyone, decide for me.

11/16/2007 7:57 AM  
Anonymous Give it up said...

This so-called and self-described "alternative proposal" is perhaps the most deceptive, self-serving and potentially dangerous scam to be seen on the south coast in some time. It is basically a one-man show [Marc McGinness] and the appalling lack of sensitivity to the underlying issues of mental illness and suicide prevention displayed by McGiness under the guise of "alternative proposal" is truly worthy of a case study all in itself. Shame on you Marc, Shame on you. Despite his efforts to intimidate, manipulate and play a shell game on elected officials NOT ONE, count them NOT ONE has bought into his bs and supported him. He has one goal in mind and it has nothing to do with preventing suicides. HIs goal is to prevent any "change" to the physical appearance of the bridge. He has ZERO expertise in mental illness, law enforcement safety techniques yet he is dangerously presenting distorted and manufactured "alternatives" which could exacerbate danger to both.

Sara, I really wish you would stop offering him free advertising for his dangerous and deceptive scam.

11/16/2007 8:59 AM  
Blogger jqb said...

They are very clear: the world is better without the people who commit suicide off the bridge. They don't value human life. They are the analogues of Bush's torturers on the left wing.

What makes you think they are left wing? And how can they be "very clear" about something they never say or imply? I you're torturing logic and good faith.

11/16/2007 10:15 AM  
Anonymous Anonymous said...

My initial "gut reaction" to the idea of a suicide barrier on Cold Springs Bridge was completely negative. In fact, I would prefer that the existing structural roadside guardrails be lowered to permit a view right out the auto window.
But The Glendon Association, our local professional suicide research, counseling, & prevention experts, have proven by research studies that suicide is an "interruptible" event. Most that commit to this "final" resolution if deterred or stopped, never again take this "final" step.
How many "gifts of life" does it take to be worthwhile?

11/16/2007 1:07 PM  
Anonymous Anonymous said...

Maybe we'll save a Vet. CBS News reported appalling statistics of the homeland carnage.

Suicide Epidemic Among Veterans Nov 13, 2007
A CBS News Investigation Uncovers A Suicide Rate For Veterans Twice That Of Other Americans
http://www.cbsnews.com/sections/eveningnews/main3420.shtml

There are already stories deflecting from this tremendous pain. It can't be that bad. Dr. Laura will tell you what the Wall Street Journal has to say.

11/16/2007 5:39 PM  
Anonymous Don Jose de la Guerra y Noreiga said...

Anonymous 5:39

Speaking of Veterans, I'm wondering when Veterans for Peace is going to give an accounting of how they have been spending the donations coming in for Arlington West? What's hanging this inquiry up? They get free beach space and might be commiting a fraud. Let's find out the truth.

I don't know, but maybe some of the real soldier's depression and subsequent suicide comes from the sham of Veterans for Peace. Or maybe it's code pink.

11/16/2007 7:36 PM  
Anonymous sa1 said...

I'm with Observer on this one.

Funny how we choose when to care about a single life...

Usually it's right about death time... the medical communities last chance to suck any remaining assets out of a person before they're allowed to expire.

11/16/2007 7:41 PM  
Anonymous sa1 said...

You know, the more I think about it, the more that front page story about the homeless Meth Ho pisses me off. $100K a year in medical costs for the county to keep her alive. 20 years she bummed off the system cuz she couldn't handle the real world. If we point ten of those to the bridge, maybe we'd have the funds to treat those who actually want and seek out help.

The world's full of tough choices.

Lining the sidewalks with fluffy cotton so we all don't get hurt isn't one of them

11/16/2007 7:52 PM  
Anonymous vocevida said...

The alternative proposal offered by Friends of the Bridge is based on solid evidence and "best practice" for a transportation agency. I have carefully read it and compared it to the CalTrans proposal, and I recommend that people who really care about the problem of suicidal behavior on the bridge not jump to hasty conclusions until they have done so.

Read. Think. Speak. Listen.

Which of the two proposals merely diverts suicidal behavior to someplace else in the community? Which does not?

Which of the two proposals diverts $1,000,000 from limited funds needed to address traffic safety problems elsewhere along Highway 154 that are more far more severe in terms the numbers of people who are killed and injured every year? Which does not?

These and related questions have clear answers. But to find them, one must do the reading and thinking that are required.

Do people who are unwilling or unable to do this really deserve to be listened to?

Read. Think. Speak. Listen.

11/17/2007 3:46 PM  
Anonymous Anonymous said...

Give it up, give it up.

McGinnis is just one of the great many of us Friends of the Bridge who, like him, care deeply about the problem of the suicides that occur from time to time on the Cold Spring Bridge and who are determined to make sure that the problem is addressed by effective rather than ineffective measures. Check around. Most people who are informed about this issue-- the number is growing steadily-- will tell you which proposal makes more sense.

Many of us who recognize that the CalTrans proposal is the wrong response are conservatives who are rightfully suspicious of folks who seem to believe that we need to be saved from ourselves and need to waste our resources save them from having more important things to do.

Many of us remember that CalTrans tried to sell us on its idea that we needed an elevated freeway running through the heart of the Santa Barbara community? I have to guess, give it aup, that you would have given up and given in to CalTrans then, and where we the rest of us be now if people like McGinnis were not willing to take CalTrans on and turn it around.

The point is that we many Friends of the Bridge are confident that most people in the community will come to support our proposal and reject what CalTrans would rather do.

Conservatives, like liberals, know that if the people have what it takes to lead, the leaders will eventually want to join in.

Finally, please try to curb your awful name-calling and venomous caterwalling.

Just give it up.

11/17/2007 4:42 PM  
Anonymous Anonymous said...

Funny how the article calls a barrier both a quick fix and a boondoggle. Which is it? That is what Reading and Thinking brought to me.

Another bit of thinking... where is the Glasgow study that supports *any* of the measures proposed in this study? There is none. And Glasgow is quite hard to pin down on what he means.... he freely says that barriers might work fine, just his study was unable to verify that.

Read. Think. Ask Questions.

There is not one scintilla of evidence that a person prevented form suicide at the Cold Springs Bridge will go on to commit suicide elsewhere. Not one scintilla.

Read, think, ask questions.

There is substantial evidence based on over 500 interviews of suicidal folks saved on the Golden Gate Bridge that they did not go on to commit suicide elsewhere.

And of course: with all this ballyhoo that McGinnes is making, suicidal folks will still just avoid the Cold Springs Bridge.

The cost of the McGinnes proposal will be far, far higher when you add it up over the life of the proposed barrier, 50 years or so.

If you ask any hard questions of McGinnes, he accuses you of name-calling and caterwalling. His skin is about one molecule thick.

11/18/2007 7:09 PM  
Anonymous vocevida said...

Anonymous 11/19 7:09:

A " quick fix" that does not work is a boondoggle. Read what the former director the the New York State Office of Mental Health said about this. He is the one who recommended against an attempt at a quick fix in the form of installing bridge barriers in favor of the kind of approach set forth in the alternative proposal presented by Friends of the Bridge.

Glasgow's work clearly shows that costly and defacing bridge barriers so not save lives, and it should be obvious that it would be a costly mistake to install barriers on the Cold Spring Bridge where the incidence of suicide is very, low (1 per year on average-- a very small fraction of the number of suicides that occur elsewhere and by other means in the county every year.

The evidence that barriers do not save lives (as determined by no observable change in the suicide rate in the community where the bridge is located) is evidence that indicates that if a person is prevented by barriers from jumping from a bridge will simply be diverted to commit suicide elsewhere and/or by other means. Follow?

Next in order:

What occurs on the urban Golden Gate Bridge has nothing to do with what goes on at our rural Cold Spring Bridge. Here is why: As Seiden well understood but did not reveal in his infamous "study" of those who were dissuaded from jumping from the Golden Gate Bridge, many if not most of the 551 people he interviewed were what he called "pseudo suicides" who never intended to commit suicide in the first place. Ask the folks at CalTrans and/or Glendon to share the information about this that Friends of the Bridge has provided to them. Hint: What he revealed that he knew in the Time magazine article.

I doubt McGinnis reads this stuff, and I know it was not he who made the caterwalling comment. It was me, a person far more conservative than him. On this issue, most conservatives that I know agree with most of the the liberals that I know in concluding that the barriers proposal is a wasteful and misguided boondoogle which should be withdrawn in favor the no-barriers altenative proposal based on what the New York Bridge Authority folks have done on their bridges.

Your questions are not hard ones. The answers to them are easily provided based upon solid evidence. Whether or not you are satisfied by them is your business of course, but you might want to try to avoid embarrassing yourself by what you write here.

11/18/2007 8:33 PM  
Anonymous Anonymous said...

8:33pm and where has Glasgow proven that any of these measures now proposed actually work? He hasn't.

There is one very low standard of evidence for things this group wants to do, and another much higher standard for what they oppose.

You are quite wrong about Glasgow. He freely admits he has in no way proven that barriers are ineffective. He has been unable to prove that they are effective, which is a very different thing. Ask him.

Show me the study that proves that there is a statistically proven diversion of suicides from a barrier. You can't because no such study exists. The diversion is just a conjecture, no more.

Sure, Seiden is imperfect. But his work soars above the sophistry of the anti-barrier folks here. No amount of words alone can disprove his core conjecture; if you want to disprove his work, you better go through and do a better study where you make numerical estimates of displacement/diversion and whatever other effects you believe afflict his work.

For example, you could do a survey that deterimines what fraction of the people in communities with bridges with barriers are even aware that the barriers exist. The anti-barrier folks have asserted that 100% of folks know that barriers exist on such bridges, and therefore 100% of suicides will be diverted. That is pretty bizarre.

And the anti-barrier folks assume that potential suicides will not be repulsed by all the measures they propose. Again, a bizarre assertion.

Time Magazine? Give me a break. Compared to refereed journals, that is about like quoting the National Enquirer.

And where is the cost estimate for all of these proposed measures, for the next 50 years? As a conservative it would seem to me you would want to know that.

11/19/2007 6:58 AM  
Anonymous Anonymous said...

The detailed report by the New York Bridge Authority is at:

.pdf of Report

It doesn't seem to offer any statistical studies of the human barrier proposed in NY or that the proponents want for the Cold Spring Bridge. There seems to be no proof at all offered that the Human Barrier would save any more lives than a physical barrier would, and certainly no proof that the overall suicide rate goes down when `Human Barriers' are deployed.

Indeed, on page 14 of the NY report, the Seiden study is lauded and accepted. The New York study raises no concerns about displacement of suicides. Displacement of suicides to a new location has been a prime argument of our local anti-barrier folks.

Many considerations went against physical barriers for the New York bridges that do not apply for Cold Spring: snow, replacement of expensive equipment already owned for maintaining the bridges, wind, and the shear length of the New York Bridges.

In addition, the New York group is copying the Golden Gate Bridge system which they laud. That is very troubling... there are still so many suicides with the Golden Gate Bridge Human Barrier system in place that a movie (The Bridge) was made about the suicides that take place even though the Human Barrier there is present.

Seems to me that the phone systems, cameras, and patrols could be deployed *in addition* to the physical barrier on Cold Spring. Belt and suspenders, physical intervention and human intervention.

11/19/2007 6:25 PM  
Anonymous Anonymous said...

Is it true that there have been 43 suicides from the rural Cold Springs Bridge over its 44 years of existence?

Don't guess. The answer is Yes.

By what kind of tortured logic could anyone in their right mind support diverting One Million Dollars to put up barriers from funds that are supposed to address traffic safety problems that kill many times more than the one person a year who chooses to jump.

Spare us from having to listen to the bit about "if it saves even one life, it is worth it." Bunk! Far more than one life a year might be saved if One Million Dollars were spent on fixing problems that kill many times one person a year.

Dio you know that determined jumpers have brought ladders to climb over barriers?

I bet that the folks in New York have wised up to Seiden's deception by now. Someone should check with them. Anyway, there is no way that one can logically compare what happens on the bridge here (very few attempts at suicide) and the Golden Gate where lots and lots of people from all over the world come to mostly to fake suicide in order to get help or to end their lives in the way they choose which is no crime.

Bridge barriers are a kind of twisted effort of big brother authorities dressed up in drag as fretting nannies to make sure that we do not remove ourselves as consumers and taxpayers ahead of what their schedules call for.

11/19/2007 9:54 PM  
Anonymous Anonymous said...

hmmm anon 9:54...another "mental health expert"--right....NOT.

You are comparing apples and oranges and trying your hardest to exploit arguments that you hope will trigger the support of south coast "progressives" who are well-meaning but who no doubt won't look much beyond your superficial platitudes.

11/20/2007 6:45 AM  
Blogger MCConfrontation said...

Let them jump. Why does Uncle Sam get to decide who jumps and who doesn't? Let's see, a million bucks for a guardrail that might not even work, or a couple hundred bucks for a couple of caltrans guys with shovels at the bottom of the ravine once a year. I'll take the latter.

Oh and it costs a whole lot more to clean up after a guy that jumps in front of a train, which happens every once in awhile around here, then it does for a guy who hangs himself, or jumps off a bridge, or OD's on pills, or straps a 60 pound divers vest to himself and paddles out to a depth of 15 feet of water in a pond in an inflatable raft and slashes the raft (like my brother in law did).

So how come we don't have military police at every train crossing keeping the loonies from kersplattering themselves in front of the Amtrak? Shoot that's happened like three times in the last year! What does it cost to clean up that mess?

Memo to Uncle Sam: stay out of my suicide plans.

11/20/2007 7:44 AM  
Anonymous Anonymous said...

9:54pm how about some proof that $1 million spent elsewhere would save many more than 50 lives over 50 years, which is what $1 million the Cold Spring barrier would save?

11/20/2007 8:30 AM  
Anonymous Anonymous said...

If you look at the New York State Bridge Authority article, you will find that they have proposed a plan that suits their needs, not that of SB. Barriers on the Five bridges in their area, would be very expensive, and would interfere with snow removal and bridge repair, creating more of a hazard. This is not true in SB. The bridges referenced in New York have pedestrian access which allows them to place the phones along the bridge itself; we cannot. If you look in the appendix of this article you will find that the costs of training a small number of authorities to monitor the bridges, installation of cameras and phone lines, would surpass the million dollars allocated to our Cold Spring Bridge.
It is also funny to me that they use this article after knocking Seiden's study; there is a whole section that references the importance of Seiden's bridge study.
I used to study under McGinnes and am appalled at his statement that "The small number of people who jumped from this bridge cannot be deemed the reason for making all of us victims of their suicide." The NYSBA article touches on media sensitivity to suicide, I suggest McGinnes and Glasgow follow this advice.
Furthermore, as a sociologist I am appauled by Glasgow's circulation of his so called study. It is not backed by the University and has not been peer reviewed or published by any acredited journal. Ironically they use his paper as the main resource for their alternate plan. I listened to 990AM on Saturday, and Glasgow sunk against authorities in suicide. I highly suggest finding it in the archives. They have no expertise or knowledge of suicide and clearly are ignoring evidence that removing means is one of the top five strategies backed nationally for suicide prevention.

PS the Brooklyn Bridge has barriers, so does the Eiffel Tower are they misguided for believing in the effectiveness of barriers.

11/20/2007 4:30 PM  
Anonymous Anonymous said...

I think MC is right in pointing out the seriousness of suicides and accidental death along the train tracks here.

The accidentals bother me the most, although MC would probably say anyone stupid enough to walk along the tracks deserves their fate. I wish we could deploy some kind of technology that senses people along the tracks and then does something... perhaps activates a sound warning near the person in question (ooo, the train track neighbors in Montecito will love that) and displays the person's location back to the train conductor. Seems to me with all the technology we have these days something like that must be possible, particularly at night when I think most trains go through and the infrared signatures of humans are pretty distinct from backround.

I disagree with MC on suicides... I think people commit suicide out of acute overreaction to problems that seem insoluble but are in fact soluble. I think society benefits a bunch (in not having lost wages and taxes of the suicide, as well as those of devastated relatives, and savings in public expenditures like morgue, police, etc) when people are prevented from committing suicide.

11/20/2007 7:09 PM  
Anonymous Anonymous said...

McConfrontation: You are right on.

I really do not want nor does anyone NEED "Cradle to Grave" care.

Kervorkian did a good service for those afraid to pull their own plug... Let's hear all the screaming at this statement.

Meantime, save the Million for something stupid in Art that the Mayor will come up with next.

11/21/2007 1:51 AM  
Anonymous Anonymous said...

When the Mayor was in Denver recently, she signed on for parking meters for people to give to the homeless--never mind that we don't have parking meters in SB. Whenever she's with her fellow mayors, Mrs. Blum just has to show her "leadership." Then she has to come back and sell it the people who live here. Didn't work with the blue line, I wonder how this one will set with the citizenry.

11/23/2007 10:49 AM  
Anonymous Anonymous said...

Hi, Garrett Glasgow here.

Let me lay out my position again, since so many people seem to be confused about it. What I did in my report was review past research on the effectiveness of suicide barriers. What I found was that there was no conclusive evidence that suicide barriers save lives (yes, that includes the Seiden study). My own statistical study used more data than any previous study, and I still couldn't find any evidence of a life-saving effect for suicide prevention barriers.

It is possible that a barrier might save some lives and we're just unable to detect it (for instance, if someone walks to a bridge, sees a barrier, and walks away and lives, we'll never know). However, we have no evidence that this is happening -- it's just as likely they walk away and commit suicide some other way.

Thus, neither I nor anyone else that has ever studied the problem has been able to find any evidence that suicide prevention barriers save lives.

Therefore, I voiced my concern that we were planning to divert $1,000,000 from the highway safety budget to a barrier, when we had no idea if it would actually make our community safer. In fact, a barrier might make us less safe by underfunding other, more effective highway safety projects.

I'd like to point out that so far very few critics have even tried to address these points. For instance, nobody has come forward and tried to explain why they think I've interpreted a study incorrectly, pointed out a flaw in my statistical study, or tried to argue that the barrier is the best use of our safety money. Instead, I'm getting ill-informed complaints that sound like they came out of some kind of political focus group.

For instance, one complaint is that I have no expertise in this area. On the contrary -- I have extensive training and expertise in social science statistics (check my vita if you like), which is exactly the topic of the studies cited in support of the barrier. In fact, one could fairly ask what statistical expertise the barrier advocates have ...

Another complaint is that I've been criticized for circulating a non-peer reviewed study. Anyone who actually understands the peer review process will find this argument absurd. Universities never "back" the research of their professors, and it is common practice to circulate working papers that have not yet been through the formal peer review process (in fact, informal circulation is often regarded as the first round of peer review). Click around on a few faculty member's websites at UCSB and you'll see a lot of them have posted "working papers" that have not yet been accepted at journals -- this is standard practice.

Further, my report is best regarded as a peer review of previous research, since I'm simply reviewing past claims and interpretations. Finally, I suspect another reason why barrier proponents want me to seek out peer review and publish in a journal before they'll address my points is because the entire formal peer review process can take more than 2 years ...

Another angle that is popping up on the blogs is that I "sunk" in a radio interview on 990AM. The radio interview was an interesting experience. I had to deal with a medical emergency at home, so I didn't get to call in until very late in the show. Arrayed against me were Tom Woodrow (the host), his father, and Dr. Firestone.

It was obvious that the host and his two pro-barrier guests were quite emotional about the issue, and it was clear that at least two of them (the host and his father, who was one of the experts) hadn't actually read my report. Dr. Firestone barely spoke, so I couldn't get a read on what she knew. I found it impossible to steer the conversation to a discussion on the facts -- instead, the conversation was me answering a question or two from the host, and then the two other guests telling emotional, irrelevant stories. I was never given the chance to ask the barrier advocates questions.

I think Tom did want to give me a fair hearing, but his own beliefs kept him from asking tough questions of the other side. Oh well. I did my best to be respectful of my critics, and even let some minor misrepresentations of my position slide by. I answered every question honestly and thought I explained my position well in the little time alloted to me. If that's "sinking," I'm happy to sink every time.

Finally, for the person who suggests my report would get a D in in any serious upper-division statistical methods course, I invite you to sit in on the serious upper-division statistical methods course I'll be teaching this spring.

11/23/2007 9:21 PM  
Anonymous Anonymous said...

Dr. Glasgow...

You say...

`nobody has...

1)` come forward and tried to explain why they think I've interpreted a study incorrectly'

I think several people have pointed out that you *completely* dismiss the 484 cases followed by Seiden of suicidal folks found on the Golden Gate Bridge who did not go on to commit suicide subsequently. You argue that *every single one* of these 484 cases is irrelevant to the argument for barriers. *Complete dismissal* of 484 cases seems unreasonable. It may be that a certain fraction were not really seriously suicidal and just wanted attention, and a certain fraction would have not gone to the Golden Gate Bridge had they known it had barriers. But you argue that *every single one* of those 484 cases is irrelevant. A serious job would entail sifting through all of Seiden's interviews, devising quantitative criteria for the evaluation of each interview, and assigning probabilities to various possibilities, like `not serious', `would have been diverted', `likely to be saved by a barrier', etc.. Further followup intereviews with those among the 484 survivors might have been helpful. You didn't do those followup studies, but you still blithely dismiss Seiden with a sentence or two, which is too extreme.

2)`pointed out a flaw in my statistical study'

It has been pointed out that your method is not sufficiently sensitive. Your statistical and systematic errors are so large you don't have the ability to make *any* significant statistical statement. It may be that you used more data than anyone else, but you unfortunately are still left you with an error so large you are unable to make any statistically significant statement.

Further, you have been extremely hesitant to clearly state your errors and sensitivity. Your errors and central values are not stated in your posted paper, and worse, you are defensive and drag your feet in addressing your sensitivity. It appears that you are either ignorant or deceitful.

3)`or tried to argue that the barrier is the best use of our safety money.'

There have been people who have pointed out that over a 50 year lifetime, the cost of the barrier is $20,000/year, and the Mental Health Budget of Santa Barbara County is $40,000,000/year, or 2000 times higher.

You have made no statistical study whatsoever of any positive alternative, like the `Human Barrier' proposed. I think you know that totally convincing statistical evidence in favor of any course of action is very hard and often difficult to accumulate.

It is much, much easier to do studies of insufficient sensitivity and then say `there is not evidence *in favor* of course X'. All such studies achieve is paralysis, but perhaps that is your goal.

11/24/2007 11:55 AM  
Anonymous Anonymous said...

Hi, Garrett Glasgow again. Let me address the 3 points raised above. To be fair, the poster above is right -- a few people have tried to argue the facts. However, they seem to be ignoring or not understanding my responses, which have been out there for at least a month.

1) As I've explained perhaps a dozen times before, the main problem with the Seiden study is *not* that the people in his study might not have been serious about suicide (although that is a problem Seiden should have at least mentioned).

The problem is his study does not actually address the concern that a barrier would just drive suicidal people elsewhere. The easiest way to see this is to imagine that we had installed suicide prevention barriers on the Golden Gate Bridge when it was built. What would the 515 people in Seiden's study have done? Would they have still gone to the Golden Gate Bridge, or would they have formed a different suicide plan? Remember, when those 515 people went to the Golden Gate Bridge it was possible to commit suicide there -- they were talked out of it, but their suicide plan was still viable. In order to believe that Seiden's study teaches us something about the effectiveness of barriers, we need to believe that these people would have only considered the Golden Gate Bridge (with barriers) in their suicide plans -- in other words, we must assume away the possibility that they would have formed a different plan once the barriers were installed.

The problem with knowing if people are serious or not, and the difference between human intervention versus a barrier are secondary concerns in light of this.

I explained this in my report two months ago, and numerous times since then on the blogs.

2) Well, that's a kick in the head. I spent about a month talking to this critic on the Independent site, trying to explain my study and basic statistics, and then I'm called "hesitant" and "reluctant." In fact, it took 3 messages from me before this critic even understood how my data differed from previous studies, so I strongly suspect this person is more interested in making baseless accusations than understanding my statistical study. It became very clear during our earlier exchange that this person doesn't understand basic statistics, yet I still treated him/her with respect and patiently explained everything multiple times. You're welcome.

On to the facts. In my statistical study I was unable to find a relationship between bridges and the suicide rate, a necessary condition for us to believe that suicide prevention barriers save lives.

One possibility is that there is some relationship, but it is so small we cannot detect it. Another possibility is that there is no relationship. My study is inconclusive on this point.

I should note that my results actually replicate the results in *every other* study that has been done in the past, including those cited by the barrier advocates. If you think there's something wrong with my results, then you think there's something wrong with *every other* statistical study done on the topic.

I should also note that I used the same basic statistical techniques as previous studies, so again, if you think there's something wrong with my statistical technique, you think there's something wrong with every previous study on the topic. There is no more error in my study than any previous work (in fact, there is less). The only real difference between my study and previous studies is in the data I used.

If my critic is truly interested in understanding the statistics behind my study, I'd be happy to meet for a cup of coffee and explain it in depth (you can wear a disguise if you like).

3) The barrier is *not* being paid for from the mental health budget, but from Caltrans highway safety funds. Taking money from guardrails and median barriers to spend on the barrier may have an adverse effect on highway safety, and even if a barrier saves some lives we might come out behind if more people wind up dying on the highways. This is a point barrier advocates have refused to even acknowledge.

It is true that I have not yet done a statistical study on the alternative proposal. I am in touch with the NYSBA and am planning to undertake this study soon. I'm also planning to update my statistical study with county level data, which will be even more precise than my current study.

11/24/2007 1:53 PM  
Anonymous Anonymous said...

Hi, Garrett Glasgow again. One comment and one question:

The comment: has anyone else noticed that the barrier advocates have (perhaps unknowingly) begun to concede my point that we don't know if a barrier will work or not? At the public meeting in July we were told it had been proven that barriers save lives. Now the advocates have backed off from that claim, and instead say things like "I think you know that totally convincing statistical evidence in favor of any course of action is very hard and often difficult to accumulate."

So we're actually agreeing on this point. It is hard (but not impossible) to find clear evidence one way or another on the effectiveness of barriers. Right now we just don't know the answer.

It doesn't necessarily follow from this that we shouldn't build the barrier. It is a perfectly valid position to say "we don't know if barriers work, but we should build them anyway just in case." I believe we need to carefully consider the tradeoffs (highway safety versus suicide prevention), but arguing in favor of the barrier is not by itself the wrong position to take.

However, we need to be honest with the public about what we actually know, and right now we don't know if barriers work. Misleading the public by claiming there is "proof" that barriers work when there is none will not help us to spend our safety money wisely.

And my question: to the individual criticizing the statistical work in my study, could you please tell us about your background in statistics?

11/24/2007 3:19 PM  
Anonymous Anonymous said...

Dr. Jamie Rotnofsky is the Executive Director of the Glendon Association. Her website clearly reflects her dedication to serious, scientific research:

http://www.askdrjamie.net/

She got her PhD from the Union Institute, an online correspondence school:

http://www.tui.edu/psyd/admissions.asp

I'm sure it's very prestigious, at least for an online correspondence school.

Ladies and gentlemen, your local suicide prevention expert.

Why are we taking these people seriously?

11/24/2007 9:47 PM  
Anonymous Anonymous said...

1)Dr. Glasgow, you did not address the issue that anon 11:55am brought up. You dismiss *all* 484 cases brought up by Seiden because you think they would have made alternate plans had barriers existed on the Golden Gate Bridge. This is not reasonable because:

a)you have not established what percentage of either the general public or potential suicides are aware of barriers when they are placed on a bridge. You assume the percentage is 100% without evidence.

b)You have not established that 100% of people in emotional distress acute enough to go to a bridge to commit suicide would consider the existence of barriers in any way.

2)Your characterization of the exchange about statistics elsewhere is not accurate. I have a Ph.D. in Statistics from Stanford. The questioner simply wanted to try to establish if your errors on the correlation coefficient of total suicides with number of bridges were small enough to see the expected correlation due to actual suicides from bridges under the assumption of no displacement. You were not forthright with a numerical reply, and you still are not forthright here. This question can be and could have been answered in one sentence.

3) If you really want to optimize the expenditures of money to achieve the most impact on folks in mental distress, why don't you focus on the $40,000,000/year spent by County Mental Health? If you could improve the expenditures there by just 1%, you'd have twice as much impact as the barrier expenditure.

I don't recall that you have done any comparative study of Caltrans needs for road safety. Caltrans itself has a vigourous internal program for such comparative analyses, which you don't seem to have studied. How do you know whether or not Cold Spring Barrier money would have more impact if spent elsewhere? Caltrans itself has chosen to spend the money on the Cold Spring Bridge.

There is strong (but not utterly convincing) evidence that barriers save lives... the Seiden study. It is not perfect, as you point out, but even with its flaws, it is sufficient to justify barriers.

Wouldn't you rather err on the side of saving lives? In a different situation, say, the emission of toxic chemicals into the environment, would you demand utterly convincing proof that the toxic chemicals cause cancer or other harm before regulating or prohibiting their emission?

As for the `Human Barrier', they use that on the Golden Gate Bridge. The result? So many suicides remain that a major documentary was made about the very frequent (several per month) suicides there.

11/24/2007 10:21 PM  
Anonymous Anonymous said...

Hi, Garrett Glasgow here.

To the "PhD in Statistics from Stanford", please give it up.

You are not a PhD in statistics. You are the same person as before, asking the same questions as before because you didn't like the answers I gave you the first half-dozen times.

You are a tedious internet crank, and I'm done wasting my time with you.

For the rest of you, if anyone out there has actual questions or comments related to my study or any of the other evidence related to suicide prevention barriers, please feel free to contact me at my UCSB office, either by phone or via email.

11/25/2007 5:27 PM  
Anonymous Anonymous said...

Hi, Garrett Glasgow again.

While I've answered all the questions put to me by anon 10:21pm already, it occurs to me that perhaps I haven't explained the answer to one question clearly in the past.

The question is: are the data I'm using actually precise enough to detect changes in the suicide rate related to bridge suicides? I answered this question in a really technical way before, so I'll give a simpler answer here.

The short answer is "yes." I just did some simple t-tests, and I have enough observations to detect a change in the suicide rate if even 1% of suicides by jumping are averted. I did this by simply creating a new variable where I subtracted that number of suicides from the total number of suicides, calculated that new suicide rate, and compared it to the real suicide rate with a t-test. This is the advantage my study has over others -- with so many more observations (1,325), I can detect much smaller effects on the suicide rate than previous studies.

So, my study should be able to detect any reasonable sized increase in the suicide rate due to bridges. The fact that I was unable to do this means my study does not provide evidence that suicide prevention barriers would save lives. This doesn't *prove* that barriers don't work, but it is troubling that even with a great deal of data we can't detect any relationship between bridges and suicides.

11/25/2007 6:16 PM  
Blogger Sara De la Guerra said...

Garrett -- thank you for being so patient with our readers questions. It amazes me that people, like some detective in an interview room, think they will get a different answer if they ask the question in a slightly different way.

In any event, I find it fascinating that this post gets so many comments as if it is some sort of ad hoc peer review....clearly not but I hope you can see what I am saying.

Thanks again for your clear answers and willingness to answer questions from our readers.

11/25/2007 6:26 PM  
Anonymous Anonymous said...

Hi Sara, Garrett Glasgow here. Sorry, I kind of lost my temper a few posts ago -- apologies all around.

I don't mind sticking around to explain things -- this is a tricky, complicated issue, and if I can clear things up for a few people, I'm happy to do it.

I did realize I should answer one other question that came up about how Caltrans allocates safety funding. Basically what Caltrans does is assign each project a "safety index", which is a cost/benefit calculation comparing the cost of the project to an estimate of lives saved and property damage averted.

My concern with the safety index calculation for the barrier on the Cold Spring Bridge is that Caltrans made the calculation under the assumption that barriers would save one life per year. Under this assumption the safety index is though the roof, and the barrier project becomes what is known as a "0-1-0" project, or a "high priority spot safety improvement" project. This means it is expedited over all other safety projects.

Of course, we don't actually know if a barrier would save one lifer per year -- it is pure guesswork at this point, and you could just as easily plug other numbers into the safety index calculation that are also consistent with the data (such as 0 lives saved) that would move it out of the 0-1-0 category.

What projects Caltrans is delaying or underfunding in order to prioritize the barrier? Are there other projects that have been are proven to save lives, but have a lower safety index than the barrier because of the one life a year assumption and have thus been delayed? I'll try to find out from Caltrans over the next couple of weeks ...

11/25/2007 7:16 PM  
Anonymous tedious internet crank said...

Sara - you are wrong here. I'm not asking the same, slightly different questions thinking I'll get a different answer. I keep asking in the hope that Garrett Glasgow will actually answer, and he has not.

He has nowhere stated his correlation coefficient and its error, which is basic to science and to evaluation of his result.
He has not stated the expected correlation coefficient in total suicides from bridge suicides alone.

The t-statistic he mentions is part of the story. But the t-statistic only involves the ratio of the coefficient to its statistical error, and not actual value of the correlation coefficient.

I've never made a personal attack on Prof. Glasgow of the sort that he just did on me. That he descends to such an attack, rather than provide 3 simple numbers (a correlation coefficient, and error, and an expected correlation coefficient from bridge suicides) diminishes the seriousness of his work.

Simple random variation would require >40,000 bridge suicides to convincingly detect a 1% effect (1/sqrt(40000)=1/200=0.5% for a 2 sigma effect at 1%). There aren't >40,000 bridge suicides in the US in the time period he is using.

11/25/2007 7:47 PM  
Blogger Sara De la Guerra said...

tedious -- I didn't get the sense that this was one person...just a sense among skeptical people. If Garrett hasn't answered this perhaps he will -- I'm too confused by statistics to keep track. Thanks for checking in.

Garrett -- I would lose my temper too but hopefully each of us can see that the perspective is different when you have a different lens on.

11/25/2007 8:16 PM  
Anonymous Garrett Glasgow said...

Hi, Garrett Glasgow here.

Kind of funny that the person who called me "ignorant" and "deceitful" is now complaining about personal attacks. However, they're absolutely right that I shouldn't make assumptions about people, and I apologize for any offense I caused.

Anyway, here are the numbers you were asking for:

Defining bridges as over 30m high with pedestrian access, using heteroskedasticity corrected standard errors, and a lagged dependent variable and the unemployment rate as control variables, I get a regression coefficient on bridges of -0.025 with a standard error of 0.014 when I use the overall suicide rate as the dependent variable, and a coefficient of -0.003 with a standard error of 0.001 when I use the suicide rate by jumping as the dependent variable. Note that second regression does not distinguish bridge suicides from other jumping suicides, so you could apply some kind of correction to that number if you like. I can email you my data if you'd like to double check these results yourself.

I should also point out that your sample size calculation is wrong. I think you're confusing effect size with significance level.

One easy way to see your formula is wrong is that it does not account for the standard deviation of the variable you are examining. This matters because it is harder to detect meaningful changes in variables subject to more random variation (with larger standard deviations) than in variables subject to less random variation (with smaller standard deviations). To see this, take it to the extreme -- we'd only need a couple of observations to detect a change due to some treatment in a variable with no random variation (a standard deviation of 0), since any change would be due to the treatment.

Of course, the easiest way to see your calculation is incorrect is to realize that that results of your calculation are absurd. How can it be we would need 40,000 observations to say something meaningful about a variable? Most public opinion polls have sample sizes of around 1500. Most psychology and economics experiments have fewer than 100 observations, and they certainly don't have effect sizes of 50%. Yet they still produce meaningful results. I'd recommend grabbing a good textbook with a chapter on power calculation if you want to read more on this.

In other words, we don't need to observe 40,000 bridge suicides in order to say something meaningful about the relationship between bridges and suicides.

11/26/2007 12:43 AM  
Anonymous Anonymous said...

Sara--- since Mr. G clearly, in true academic fashion, has to have the last word---could you send him instructions on how to set up his own blog? Then he could tell the world, ad infinitum, about his superior knowledge and understanding of mental illness. until then, YAWN

11/26/2007 7:38 AM  
Anonymous tedious internet crank said...

Those are two out of three of the numbers needed... still need the expected correlation coefficient from known bridge suicides. That should be equal to the number of bridge suicides 100,000 per year divided by the mean number of bridges in your geographic subdivision (state, I believe).

The Poisson standard deviation on N events is sqrt(N). When talking about total rate of suicide (and I am) I know of no way to achieve a smaller error. You have said you would be sensitive to 1% of total bridge suicides being averted. That is a very strong statement, and to prove that at 2 standard deviations, you need a statistical error of at most 0.5% (and most people would say you would need to prove it at 3 or 4 standard devations).

Public opinion polls with 1500 respondents generally quote a margine of error of at least 1/sqrt(1500) or 2.6%. Usually bigger because there are errors in the apportionment of respondents relative to the intended population.

I never said such a public opinion poll could not yield meaningful results. But I would say the only meaningful results are bigger than 2 or 3 times 2.6%.

I've never said you need 40,000 bridge suicide to get meaniningful results in a general way. What I do maintain is that if you say you are sensitive to diversion of 1% of bridge suicides, you need to study at least 40,000 bridge suicides.

By the way, the variance on most fatalities exceeds that of a Poisson distribution. Statistical studies that Caltrans performs of highway fatalities actually account for this inflation of the variance. George Polya, who taught at my alma mater, was a bit of an expert on this.

I see now your reply about Caltrans safety study. Nice that you mention that if even one fatality per year is prevented by the barrier, this project gets a very high safety index. What continues to surprise me is that you argue that the Seiden study is *completely* wrong, and the true number of saved lives might be *0*.
As I will continue to argue, you are being extreme in your contention. Maybe the true number of saved lives is 0.7/year or 0.5/year or 0.3/year... you don't seriously consider such possibilities. You argue the extreme and complete repudiation of Seiden, which is unreasonable.

As for personal attacks, being called ignorant is not an attack. I am definitely ignorant of most of the knowledge in the world. I can only speak 3 of the 1000's of languages in the world, and I don't know anything about Thai poetry. Being ignorant is nothing to get offended over.

11/26/2007 8:57 AM  
Anonymous tedious internet crank said...

I forgot one thing... Prof. Glasgow, you say:

"and a coefficient of -0.003 with a standard error of 0.001 when I use the suicide rate by jumping as the dependent variable"

Footnote 14 on page 6 of your paper says:

"Note that there is a positive relationship between the number of bridges in a state and the suicide rate by jumping."

Which is it... negative as you said in your 12:43am post, or positive as you say in footnote 14 on page 6 of your paper?

Sorry Sara, if such questions cause you to lose your temper. After all, I'm a tedious internet crank.

11/26/2007 9:02 AM  
Anonymous sevendolphins said...

Dr. Glasgow referred me to here from the Indy discussion.

The discussion here is so long I can't really absorb it.

But it would be great, Dr. Glasgow, if you could e-mail me your data. My username is sevendolphins and my e-mail is at gmail.com.

Thanks!

11/26/2007 4:47 PM  
Anonymous Garrett Glasgow said...

For those that want my data, I'll try to get it into a user-friendly format and send it out in the next couple of weeks. Email me if I forget (we're in the middle of interviewing for a new position in my department, and things are quite hectic).

For those bored by this, stop reading (or ask people to stop asking me the same questions over and over and over ...)

I got a positive coefficient on the regression of bridges on jumping suicides with a simple regression, but it became negative once I included control variables. That at least is a legitimate question.

To my persistent critic:

Let's stop playing games. You don't really know that much about statistics.

The standard level of significance in the social sciences is 5%, or 1.96 standard deviations from 0. Nobody is arguing that we really need to be 3 or 4 standard deviations from 0 (and nobody calls the level of significance "sigma" either).

You have also totally ignored the fact that I pointed out that your power calculation was wrong, and in fact you even repeated the error. You probably didn't even read what I wrote about it.

You still haven't understood my criticism of Seiden, probably because you haven't read it.

It is abundantly clear that your goal here is to be annoying rather than have a meaningful exchange, so I'll stop here.

11/26/2007 7:49 PM  
Anonymous tedious internet crank said...

Very interesting reply about the regression coefficient of jumping suicides... why did your report not mention this detail? What is your interpertation of how presence of bridges inhibits people from jumping to their deaths? Why don't you evaluate the regression coefficient with *only* bridge suicides? If that came out negative what would your interpretation be?

I keep asking some questions over and over again because you do not answer them. The most important unanswered question: What is the expected regression coefficient of bridge suicides with number of bridges if there is no redirection? This is the bridge jumping rate per 100,000 per year divided by the suitably weighted mean number of bridges per state.

I've asked that many times and you haven't answered. One simple number, and you've poured out hundreds of words without giving that number.

Nixon had a phrase for this: sandbag `em. And you are sandbagging, Prof. Glasgow.

As for understanding of statistics, what a weird answer. You give vague, qualitative statements about irrelevant issues like public opinion polls and social science standards, but you avoid a direct formula and crisp numerical reply. Who the frig cares whether some subfields say `sigma' or not, the concept is clear. Feynman would box your ears and call you a twit. Indeed some subfields demand 5 and even 6 sigma (yes, some subfields use that term) before they accept a new result. Exceptional claims sometimes require exceptional evidence.

I very much understand your criticism of Seiden: that you think potential suicides would have taken barriers on the GGB into account in making their suicide plans, and redirect themselves away from the GGB. My criticism that you sandbag in response to: 1)Not everyone is aware of barriers on bridges; 2)Suicide is not always (not even usually) a rational process where people logically weigh the factors. So I don't accept your contention that Seiden is vitiated. I do accept that your concern might mean 10% or 40% or some other fraction of potental bridge suicides might be displaced, but 100% is too extreme. You don't provide any quantitative support for your position that Seiden is 100% wrong.

Look, Prof. Glasgow, lives are at stake here. Maximum thoughtfulness and also erring on the side of saving lives is called for. I am trying to save lives by persistently asking questions, to probe your work. If you find it annoying, well, grow up and get used to the realities of skeptical inquiry, and the persistence of people whose goal is to save lives.

Your flip dismissals disrespect the whole idea of saving lives and thus diminish your work. Thank goodness for the slow, careful, persistent, and superior work of Seiden and Caltrans: superior to your work, I mean.

If you would quit sandbagging and actually answer questions, I'm completely open to changing my assessment of your work. Maybe you are right and Caltrans is wrong. But since you sandbag, it is not possible to make that judgement.

11/26/2007 11:26 PM  
Anonymous Garrett Glasgow said...

To my anonymous critic:

If you truly believe what you're saying, you owe it to this community to stand up and be counted. If you've really spotted something wrong with my study you need to do more than just post anonymously on a blog.

As you said, lives are at stake.

I stand behind my work. Do you? If so, tell us your name.

The fact that you're unwilling to sign your name to what you say speaks volumes about your credibility.

11/27/2007 11:42 PM  
Anonymous Marc McGinnes said...

Tedious Crank, if you are going to invoke the name of my friend Richard Feynman, now deceased, in your twitish nit-picking with the point that Professor Glasgow has clearly made (there is no evidence that placing so-called suicide prevention barriers on bridges saves any lives), you ought to know that he would now have moved beyond the narrow place where you have gotten yourself so fixated.

Now that a no-barriers alternative has been proposed to address the problem of the infrequent suicidal behavior on the bridge without defacing its famed grace and beauty (Richard truly cherished beauty in all its forms), Richard would now be asking questions like these:

"How many people have jumped from the bridge over its lifetime?" Upon learning that one person a year on average over a period of 44 years have done so, Richard would ask:

"How many people commit suicide every year in CalTrans district where the bridge is located?" Upon learning that nearly 170 suicides take place in the district every year on average, Richard would ask:

"How many people are killed in traffic collisions in the district every year?" Upon learning that the answer is over 180 deaths per year, Richard would ask:

"Where is CalTrans getting the money to make it look like they are trying to save one life a year by putting barriers on this beautiful bridge?" Upon learning that CalTrans proposes to divert $1,000,000 for this purpose from funds budgeted for Traffic Safety-Collision Reduction purposes, Richard would shake his head and exclaim, "What?! Are they nuts?!"

Richard was a very practical thinker and problem-solver at the same time that he was engaged in this work theoretical physicist for which he won the Nobel Prize. As such, he would be a Friend of the Bridge if he were alive today, calling on CalTrans steer the better course set out in the no-barriers alternative project they have proposed.

11/28/2007 6:06 AM  
Anonymous tedious internet crank said...

Prof. Glasgow... scientific reviews are anonymous. One thing I've learned from you is that in Political Science it is not content but who says it that matters. One reason I like anonymous blogging is that it bleaches away all silly and useless credibility issues of `who said it'. What matters are the facts themselves.

BTW, Sara here is quite devoted to anonymity, for good reason. Your statement skirts her rules, actually. I think it was you who came to Sara's blog where everyone knows anonymity is a fundamental rule and started posting.

Prof. McGinnes, perhaps you could let Feynman make his own statements about the Cold Spring Bridge. I'm not sure he'd be comfortable with you making his statements without consulting him first.

But as for your arguments... Santa Barbara County spends $40,000,000/year on mental health services, according to the County Budget. Suppose those 170 suicides you mention in the Caltrans district all occur in Santa Barbara County (actually, the Caltrans district is rather larger, and includes SLO County and perhaps slices of other counties). Works out to $235,294 per year per suicide. The bridge barrier? Assuming a 50 year life, $20,000/year/suicide. 10 times cheaper.

As for traffic deaths, as Prof. Glasgow said, Caltrans already has conducted an internal cost/benefit analysis. The Cold Springs project has easily the best cost/benefit of any of them!

Now you say the Cold Springs project won't save as many lives as Caltrans says. You say it won't even prevent 0.0001% of the suicides on the Cold Spring Bridge. I think that is unreasonable. I bet it will prevent 30% at least. Even with that 30% factored in, I believe the Cold Springs project is still has the highest cost/benefit of any Caltrans Safety project.

And BTW, don't you think displacement and substitution will afflict other Caltrans projects? A drunken driver who doesn't perish on dangerous curve A will just go and perish on dangerous curve B. I think Prof. Glasgow should apply his analysis to *all* Caltrans safety projects before asserting that the Cold Springs one is particularly faulty.

11/28/2007 7:40 AM  
Anonymous tedious internet crank said...

Prof. McGinnes... two small but significant points.

I didn't clear my comment about Feynman boxing Prof. Glasgow's ears for being a bit pompous over the use of the word `sigma' with Feynman. Actually, I never even saw Feynman box anybody's ears, although in talks (yes, I too was at Caltech for a while) he could be quite unforgiving to those who hid behind language. His books document this... `Now do you think the answer will change if you call it sigma, rigma, dogma, or smegma?' That is more like what he would say.

And as for the $1,000,000 cost. I think about $300,000 of that is purely for aesthetics. Which do you care more about? If you want to use money efficiently, let's save $300,000 by letting there be an ugly barrier. But spending the $300,000 will get an aesthetically pleasing barrier.

And what is the cost of the Human Barrier, over the lifetime of the physical barrier, about 50 years? I bet it is somewhat greater than $1,000,000. Would be nice if your group gave that figure... the cost over 50 years of the human barrier. Do you really think Feynman would have like 24/7 closed circuit monitoring of the bridge? I don't want a film record of my totally safe model glider launches from that bridge.

11/28/2007 8:19 AM  
Anonymous sa1 said...

"`Now do you think the answer will change if you call it sigma, rigma, dogma, or smegma?' "

Ok, I nominate that for the Nobel Prize for Literature in the funny, if distastefull pseudo-alliteration subcatagory.

That would be so cool to have a friend like Feynman...

It's a crime that we hear more about Britney, Paris, gangbanging rapper thugs, crooked Pols and Coporate rats then we hear from great thinkers like Feynman was.

11/28/2007 10:15 AM  
Anonymous Garrett Glasgow said...

When I saw that the public was being misled about what we know about suicide prevention barriers, I stepped forward and pointed it out, because lives are at stake. I put my name on my work and made myself available for questions, because I stand behind my work.

Now we have you, someone who claims to be an expert in statistics who has caught me "sandbagging", being "ignorant", and being "deceitful" on an issue where lives are at stake.

And your only response to this obvious danger to our community is to post anonymously on a blog?

You are a fraud. It's obvious that you're pretending to be two different people (the "scientific review" line is identical to the line used by sevendolphins on the Independent site). You claim to have a PhD in statistics, yet you are unfamiliar with even basic statistical formulas and terminology.

You're not remaining anonymous to preserve the scientific integrity of "peer-review through blog trolling." You're remaining anonymous because you are a fraud.

11/28/2007 11:40 AM  
Anonymous tedious internet crank said...

Prof Glasgow... sandbagging yes, the reason being: what is the expected regression coefficient from known bridge suicides? I've asked you that 10 or 20 times, but somehow you never give that simple number. Why not?

But it was *either* ignorant *or* deceitful, not both. You pick, and there is nothing perjorative in my opinion about being ignorant. The ignorance refers specifically to what I think is your absence of evaluation of your sensitivity. There is more than just statistical significance that is important here, there is also the errors (both statistical and systematic) relative to some measure of an anticipated effect. I'm not sure you understand that point, so, you might well be ignorant. If you are not ignorant, well, then you know what I think the other choice is. My diagnosis is the result of asking you and reading sevendolphins interchange with you.

Sure, you're right, I'm only participating in blog discussions. That is all I care about. You are a genius there! Brilliant!

I don't know sevendolphins. There is another possibility: many people know that scientific review is anonymous, because it is.

Please post a formula sometime. You haven't so far. Then we can talk about it.

I've never claimed to be your peer, I guess that is your contention.

But I must say, please speak for yourself when you consider as your peers people who are trolls in blogs. I'm not with you there.

I have a lot of respect for your work. But, I think, you haven't really succeded in getting the sensitivity you hoped for. You had a good idea, and it didn't really work out. No dishonor or embarassment in that. But I don't understand why you aren't more forthright about probing the possible shortcomings of your own work.

11/28/2007 6:34 PM  
Anonymous Garrett Glasgow said...

This discussion has far outlived its usefulness. Thank you for your interest, and I look forward to seeing your work in print.

11/29/2007 9:29 AM  
Anonymous tedious internet crank said...

Yes, Prof. Glasgow: your definition of `outlived its usefulness' is `I don't want to answer serious questions about my work.'

Nonetheless, thank you for the several answers you did provide.

11/29/2007 6:11 PM  
Anonymous Garrett Glasgow said...

By the way, all this persistent critic is trying to say is that my statistical results do not rule out the possibility that barriers might save lives. This is true -- that is why in my report I say "there is no evidence that barriers save lives" rather than "barriers don't work." I've explained this explicitly to this person multiple times, both on this site and elsewhere. Yet here we are two months later still talking about the same thing.

I think I've done enough to explain the state of the evidence regarding suicide prevention barriers. Anyone who doesn't get it by now doesn't want to.

11/29/2007 6:31 PM  
Blogger Sara De la Guerra said...

53 comments later -- I don't blame Dr. Glasgow for wanting to move on. Please note that I moderated Tedious' comment at the same time. They had no knowledge of each other's last comments.

11/29/2007 7:22 PM  
Anonymous Tedious Internet Crank said...

Prof. Glasgow doesn't get it right, and that why this has grown to 53 comments.

I understand quite well his point that `Prof. Glasgow was unable to find evidence in his study that barriers save lives.'

I would add `Prof. Glasgow's methodology unfortunately left him with such poor sensitivity that he was never able to tell whether or not barriers saved lives.'

It is a basic criterion of science that you must establish sensitivity prior to making a judgement on an issue. For example, suppose I doubted the existence of left handed people, and tested the first 5 people I met, and they were all right handed. I'd say, `No evidence for left handed people'. But it would be glaringly obvious that I didn't test enough people.

That is what Prof. Glasgow is saying about barriers.... he didn't achieve sufficient sensitivity to make a significant test of his conjectures about displacement and substitution of method leading to no change in the total suicide rate. He tried real hard, and looked at more data than anyone ever did previously. He deserves an A for effort. But still, in the end, his sensitivity was not sufficient to test his hypothesis. It was a noble failure.

BTW, as far as I can tell, he has never gotten data on the actual number of suicides by jumping from bridges, on a state by state basis, which is how he conducts his analysis. He has the total suicides by falling on a state by state basis, but not the specific number by jumping from bridges. A bit odd because the whole point is to study suicides from bridge jumping.

However, he actually makes a much stronger statement when he says `there is no evidence that barriers save lives', because that is a judgement not only based on his own direct analysis, but all other work like that of Seiden. I take exception to this stronger statement. Seiden is most certainly evidence that barriers save lives. I actually agree with Prof. Glasgow that Seiden is imperfect, but I don't agree that the imperfection is so important that it degrades Seiden to the point where Seiden is `no evidence.'

11/29/2007 10:51 PM  
Anonymous Anonymous said...

Maybe the Crank is actually poor Seiden himself, the "psuedo scientist" who chose to conceal the "psuedo suicide" factor in his infamously flawed study.

11/30/2007 8:59 AM  
Anonymous Old Business said...

The documentary "The Bridge" is airing on IFC (Cox Channel 201) @ 11:30 P.M. tonight. A must see for this topic.

2/04/2008 8:07 PM  
Blogger Eric said...

Now that this Cold Spring bridge barrier issue was featured in the LA Times yesterday, you have awoken an opinion in the LA area and I feel a number of key points should be made.

My name is Eric and as a bridge buff, I am working on a project to document the highest bridges of the world and North America. While I am no expert on suicide I do know more about high bridges than anyone out there.

The first issue that should be cleared up relates to all the back and forth with bridge statistics. The Golden Gate bridge and the bridges over the Hudson in New York have nothing in common with each other or the Cold Spring arch. 90 percent of the people who commit suicide from the Golden Gate are from the 6 million plus greater San Franciso Bay area. Given the location of the Cold Spring bridge off of route 101 and its regional popularity, it is unlikely that any of the 44 known jumpers of the bridge lived outside of the greater Santa Barbara County area which has a population of just 400,000. Considering the suicide rate of around 25 people per year in Santa Barbara, that means just 1 in 25 or 4 percent of the suicides in the County are on the Cold Spring bridge.

Before adding any more statistics about bridge related suicide lets look at the barrier itself. It should be stated up front that there can be no doubt if a barrier is built on the Cold Spring bridge, it will absolutely reduce the suicide rate to zero. No one seemed to mention Toronto's famous Prince Edward or Bloor Street Viaduct which was second to the Golden Gate bridge in the number of annual suicides. For every 2 people that committed suicide off the Golden Gate bridge, 1 committed suicide off the Bloor Street Viaduct. That was a rate of one person a month versus Golden Gate's once every 2 weeks. So in In 2003, a suicide barrier called the Luminous Veil was installed. Designed by University of Waterloo Professor Derek Revington, the 16 foot (5 meter) high frame of angled steel rods solved the suicide problem but resulted in a visual barrier that is ultimately still a fence even if it is an elegant and aesthetically pleasing one. You can see images of it by typing in "Luminous Veil Bridge" in a search engine.

So the real debate is not whether a barrier will prevent suicides at the Cold Spring bridge - of course it will just as it has done in Canada. The question is whether, as someone named Garrett stated, "there is a distinction between preventing suicides and preventing suicides at a particular location."

Proponents of the barrier want to be able to sleep at night, to feel better about themselves that they "care" about reducing the Santa Barbara area suicide rate. The problem is that 96 percent of those who committed suicide in the region did not even think of using the bridge. The idea that you can prevent suicide by taking away one of the tools to do it is like telling obese people in Santa Barbara that Pizza has been banned. They would just continue to eat hamburgers, french fries, Doritos, cookies and a dozen other forms of junk food. Therapy from others or from yourself is the only way to reduce the suicide rate in any community just as it is the only way to lose weight. Most people who commit suicide do it in private so the bridge barrier is only going to deflect someone away from the bridge and into using a gun, pills or a dozen other ways that the other 96 percent of Santa Barbarans used to kill themselves in private during the same period the bridge was open.

So the real reason the barrier is being proposed is to make those of us who are alive feel good about ourselves for "trying" to prevent suicides at a known location. The problem is those 4 percent we deflect away from the bridge are going to succeed anyway by joining in on one of the much more common methods the other 96 percent are using and so a barrier will not have actually saved any lives. Since the total number of suicides in the region with or without the barrier will always be a hidden statistic, proponents of the barrier can then falsely take credit by telling themselves they have reduced the suicide rate by an imaginary 4 percent since there will never be a known stat on who decided to not commit suicide away from the bridge whose first choice would have been to do it on the bridge. If the 44 successful jumpers could speak from the hereafter I'm sure all of them would say they would have done it anyway by some other means. There is evidence that Toronto's Luminous Veil simply pushed the suicide jumpers over to the nearby Leaside bridge or back into the privacy of their homes.

The Golden Gate is a whole different bridge with different circumstances but a barrier would be an even worse idea than on the Cold Spring bridge. I'd love to explain further on that one but for now this thread is focused more on the Cold Spring bridge. It would be great to hear some opinions since the thread has been dormant for a while.

Eric

4/22/2008 4:19 PM  
Anonymous Garrett Glasgow said...

Hi Eric,

There is a new thread on this topic on this blog -- check the front page. I'd be very interested to hear your comments on the latest developments.

Oh, and you're right about most Cold Spring Bridge suicides being local -- I have the exact percentage in one of my reports, but as I recall it's about 85-90%.

5/13/2008 1:22 AM  

Post a Comment

<< Home