Monday, May 13, 2019

Researchers often report findings of no effect as if they meant something

I have been going on about this for decades so it is good to see it systematically surveyed in a major medical journal.  If medical researchers often manipulate their results it clearly subjects the results of climate research to similar doubt. In the study below, positive spin of statistically nonsignificant primary outcomes was found in 57% of abstracts and 67% of main text of the published articles.

Level and Prevalence of Spin in Published Cardiovascular Randomized Clinical Trial Reports With Statistically Nonsignificant Primary Outcomes

M. Khan et al.


Importance:  Clinical researchers are obligated to present results objectively and accurately to ensure readers are not misled. In studies in which primary end points are not statistically significant, placing a spin, defined as the manipulation of language to potentially mislead readers from the likely truth of the results, can distract the reader and lead to misinterpretation and misapplication of the findings.

Objective:  To determine the level and prevalence of spin in published reports of cardiovascular randomized clinical trial (RCT) reports.

Data Source:  MEDLINE was searched from January 1, 2015, to December 31, 2017, using the Cochrane highly sensitive search strategy.

Study Selection:  Inclusion criteria were parallel-group RCTs published from January 1, 2015, to December 31, 2017 in 1 of 6 high-impact journals (New England Journal of Medicine, The Lancet, JAMA, European Heart Journal, Circulation, and Journal of the American College of Cardiology) with primary outcomes that were not statistically significant were included in the analysis.

Data Extraction and Synthesis:  Analysis began in August 2018. Data were extracted and verified by 2 independent investigators using a standard collection form. In cases of disagreement between the 2 investigators, a third investigators served as arbitrator.

Main Outcomes and Measures:  The classifications of spin type, severity, and extent were determined according to predefined criteria. Primary clinical outcomes were divided into safety of treatment, efficacy of treatment, and both.

Results:  Of 587 studies identified, 93 RCT reports (15.8%) met inclusion criteria. Spin was identified in 53 abstracts (57%; 95% CI, 47%-67%) and 62 main texts of published articles (67%; 95% CI, 57%-75%). Ten reports (11%; 95% CI, 6%-19%) had spin in the title, 35 reports (38%; 95% CI, 28%-48%) had spin in the results section, and 50 reports (54%; 95% CI, 44%-64%) had spin in the conclusions. Among the abstracts, spin was observed in 38 results sections (41%; 95% CI, 31%-51%) and 45 conclusions sections (48%; 95% CI, 38%-58%).

Conclusions and Relevance:  This study suggests that in reports of cardiovascular RCTs with statistically nonsignificant primary outcomes, investigators often manipulate the language of the report to detract from the neutral primary outcomes. To best apply evidence to patient care, consumers of cardiovascular research should be aware that peer review does not always preclude the use of misleading language in scientific articles.


But What About Traffic? The Case for Road Pricing

Building more housing will inevitably cause more congestion. The answer is to price the space where the congestion occurs.

In my last Catalyst column, I wrote of the need for places like San Francisco to build more housing. Their shortages have created severe affordability problems, I wrote, and the answer is to relax zoning and allow more density. But these ideas often get blowback from people who live in the areas, and feel the impacts of the growth. They may be concerned about noise, crowding, and aesthetic changes, but above all, many are concerned about the added automobile congestion – hence their talking point “what about traffic?”

It’s a valid concern. Today, the densest cities  have America’s worst congestion and longest commutes, with metro San Francisco among the leaders in both. This makes sense: if more people jam into a space, and don’t significantly change their behavior (for example, by foregoing car ownership), things will naturally get more crowded. So wouldn’t increasing San Francisco’s density worsen the problem?

Well, yes – if certain policies don’t change. San Francisco and other dense cities feel crowded with cars, because the spaces where they’re stored and driven don’t require payment. As with many other “free” public goods, this has caused overuse – a tragedy of the commons. Basic economics would say that to prevent said overcrowding, the city should price these goods based on market demand, so that a scarce commodity is efficiently rationed. In San Francisco, this would mean market-based pricing for both parking spaces and roads. Let’s unpack both ideas.

First, a common complaint in San Francisco is that building more housing will worsen the on-street parking armageddon that already exists. But one reason San Francisco’s parking situation is a mess is that it’s under-priced. Most neighborhoods have no paid residential placard system – anyone can park anywhere. Even in neighborhoods that do, the placards cost a measly $136/year, and each address can purchase up to 4 placards. In both cases, this means too many people are searching for a small number of spots. When people can’t find parking in front of their homes, they waste time and add to congestion by circling to look for it.

Donald Shoup, a renowned parking guru, and professor in the Department of Public Planning at UCLA, has called for San Francisco “to charge drivers what parking is worth.” Residents would bid with each other for annual parking placards. Winning bidders would get the space; losing bidders would settle for cheaper, off-street parking (likely causing the construction of more lots or garages); while others who are priced out would forego car ownership, freeing up street space. Those non-owners could settle for alternative modes of transportation like transit and ridesharing, which would bolster the revenue for those services, improving their coverage. Shoup has even suggested using the proceeds from the parking placard system to fund transit.

San Francisco’s second congestion problem—crowded roads—could be fixed with a similar concept: congestion charging. While there are tolls at some entry points into the city, they are minor, and driving on city streets is free. This leads to the same tragedy-of-the-commons scenario. Drivers can’t move quickly, and their lingering further worsens pollution and gridlock.

In a congestion charging system, which has been widely lauded by economists from Paul Krugman to Murray Rothbard, road use is priced dynamically to be more expensive during peak hours and less expensive off-peak.

Installing this in San Francisco, via electronic tolls and transponders, would create behavioral shifts. Traffic would spread out more across the day, as some drivers amend their habits to avoid the premium rush hour fee. Some people would stop driving, instead taking transit that, because of the reduced congestion, would have lower headways and better on-time ratings. And interesting changes may even occur in the rideshare industry. With congestion charging, an Uber or Lyft ride would be more expensive. But that could impel the companies to scale up their services to carpool (as they already have to some degree), so the higher costs of movement are shared by multiple riders.

Cities that have tried congestion charging, including London and Singapore, have seen this increase in transit usage and reduction in gridlock. In March, lawmakers agreed to try the concept in New York City. And Scott Wiener, the California state senator of SB 50 fame, wants rush-hour pricing for San Francisco.

The benefits of doing this should be intuitive: competition for road space in San Francisco is fierce because of its density, and will get fiercer with more growth. But that doesn’t have to mean “more traffic.” If parking and driving is priced according to market demand, it will shift the behavior of users and the strategy of different industries, as they respond to the price signals. The outcome will be clearer roads for San Francisco.


Shenanigans with rapid transport facts

Greenies love buses but how well can we trust what they tell us about the advantages of buses?  The following episode suggests that bus data can be as rubbery as Greenie climate data

An occasional hobby of mine is writing letters to newspapers to address the state of the world. It is a petty occupation, but it keeps my hand in at writing and occasionally I get a bit of fun in stirring the pot and generating a reaction. Recently I had a letter in the Irish Times about the BusConnects plan that generated an interesting reply.


BusConnects is a plan to upgrade Dublin's transport services by introducing dedicated bus and cycle lanes along key routes all the way from suburbs to centre. Dublin is an old city with many narrow roads, and the plan has upset local residents on affected routes who are reluctant to give up their on street parking, part of their front garden, or see full grown trees cut down, just so that passers-through can pass through more quickly.

I can understand residents' disquiet, as the changes will have considerable impact on certain streets. However as a lifelong car-avoider, I can assure you that some sort of significant change is essential and long overdue. Bus services in Dublin are very unreliable and slow, with long, meandering routes and buses competing for road space with private cars. On bad weather days when car traffic spikes it can take well over an hour to travel the short 6 or 8 kilometres from the suburbs to the centre by bus. Meanwhile cycle lanes are generally either non-existent, or are so poorly implemented as to be actually more dangerous.

Letter and Response

I read the brochure describing changes to the planned Rathfarnham to city centre route, about which a number of residents had complained in previous missives to the editor. Page 16 of the brochure stated that current journey times of up to 75 minutes would be regularised to 25 minutes, and so I wrote a letter supporting the plan, judging that a saving of up to 50 minutes per journey was worth the sacrifice of a moderate amount of on-street parking and trees. Each passenger on the 600 buses per day on that route could reclaim time up to the equivalent of a working day per week, not to mention the reduction in costs for the bus companies, freeing up vehicles to provide additional services, reduction in emissions and so on.

Rathfarnham route change summary
A response in the next edition put me straight. The writer rubbished the 50 minute reduction as nonsense, pointing out that page 311 of the main report stated the maximum journey time was only 28 minutes, and thus if I saved 50 minutes I would be achieving time travel to the extend that I might meet myself coming back out of town on a return trip.

Two Versions of the Truth

I was curious and checked the quoted report. The writer was correct. The BusConnects' own website had published one set of figures in the consultation brochure, from which I had quoted, and a different and contradictory set in the main report, from which she had quoted.

No alt text provided for this image
Rathfarnham route journey durations from the full length report.
An additional discrepancy was that the brochure stated journey times after the changes would be 25 minutes, while the full report stated that in many cases journey times were already close to half of that with savings amounting to 7 or 8 minutes per journey. The report also made clear that the gains would be made not just by changes to infrastructure along the route, but by prioritised traffic signalling for buses and the introduction of an entirely cashless fare system.

I felt a bit of a fool, but recognised a situation I have often encountered in my work over the years - an organisation publishing two conflicting versions of the truth!

The Quest for Certainty

Looking in more detail at the main report, the times quoted were not real, absolute maximums, but rather ''the average journey time per half hour over the course of a normal weekday".

Immediately I could speculate that a "normal" weekday was one that did not contain unusual delays or variations, i.e. that excluded statistical outliers. This would explain how a 'normal average maximum' could differ so much from an absolute or occasional maximum, but how could I know if this speculation was accurate?

I know from previous experience that traffic surveys are heavily dependent on the time of year and weather that occurs while they are collecting data. Having complained to a local councillor about the volume of school and college traffic 'rat running' through our quiet estate each morning and evening, I was surprised to see survey equipment set up a few months later. Unfortunately this was during the unbroken sunshine of early August, when all the schools and colleges were off, builder's holidays were in effect, and many other people were away too. Needless to say the survey found traffic levels were very light and recommended no additional measures to control speed etc.

So what time of year was the route survey done for BusConnects? Was the weather poor or fair? What other things might have had an impact, e.g. road works, strikes, or public events? What statistical methods were used to determine the 'average' or what was a 'normal' day? What exceptional 'outliers' did they find, and did they exclude these or include them? All buses have been fitted with GPS devices for some time now, so data should be readily available in huge quantity and detail.

Looking for additional explanation to understand the data I found none. There was no information about when or how the assessment of journey times had been made, no explanation of how the figure of 75 minutes in the brochure came to be, no breakout of how much of the hoped for time savings would be delivered by infrastructure changes, and how much by less disruptive changes to traffic signals and the introduction of cashless payment, and no set of base raw figures for a data geek like me to sink my teeth into.

The only other source of data I could find was the official Dublin Bus timetables, which suggested that passengers should plan for a journey time of about 48 minutes along the affected route. Now I had three differing versions of the truth!


It is poor form that the supporting information published by a state body to help persuade the public about the benefits of a major and very expensive infrastructure upgrade should be so unclear and apparently self-contradictory. Unfortunately it is little surprise.

All organisations suffer similar situations from time to time. Without a central agreed data source different teams and departments can easily use similar data sources to generate widely different figures. This means that people with an existing predisposition for or against something can chose the set of figures that they wish to believe - a recipe for conflict and a sure-fire way to discredit the data management professionals involved on either side.

When such situations arise it is important to be able to work quickly back and find the source of the differences, something that is made much easier by having a proper data infrastructure and by adhering to agreed data processes and standards.

Public bodies should also adhere to the State's own advice on data transparency - i.e. that where data can be made public, it should be made public. That way any interested party can look into the data, challenge broad assertions, answer their own questions and ultimately confirm their faith in the competence of the relevant organisations.


‘Global Warming’ Not Scary Enough, Alarmists Rebrand It ‘Climate Crisis’

Since the expressions “global warming” and “climate change” do not frighten people enough, activists are proposing a shift in language to “climate crisis” or “environmental collapse,” with the help of advertising consultants.

Neuroscience research suggests that “global warming” and “climate change” do not produce a powerful enough reaction in people, whereas “climate crisis” got “a 60 percent greater emotional response from listeners” according to a recent study.

Environmental lobbying has reportedly yielded a 15-point increase in the share of Americans who believe that climate change is a serious problem, but activists are looking for ways to boost that number still further by using more explosive language.

Enter SPARK Neuro, an advertising consulting firm that measures physiological data such as brain activity and palm sweat to quantify people’s emotional reactions to stimuli.

SPARK Neuro fixed electroencephalography (EEG) devices to the heads of 120 volunteers to gauge the electrical activity coming from their brains.

At the same time, a webcam monitored their facial expressions and sensors on their fingers recorded the sweat produced by heightened emotions.

The group, which was evenly divided among Republicans, Democrats, and Independents, listened to audio recordings of six different climate phrases.

“Global warming” and “climate change” performed the worst, beaten hands down by “climate crisis,” “environmental destruction,” “weather destabilization,” and “environmental collapse.”

According to Spencer Gerrol, CEO of SPARK Neuro, there are two probable reasons that “global warming” and “climate change” perform so poorly.

For starters, they are both neutral phrases, with nothing “inherently negative or positive” about the words themselves.

Second, people have gotten used to these expressions and they no longer pack a significant punch. Both global warming and climate change are “incredibly worn out,” Gerrol said.

Moreover, if an expression doesn’t elicit a strong emotional response in the first place, it is even more likely to wear out quickly, Gerrol said.

In its study, Spark said it was looking for a “sweet spot” that provoked a response but did not backfire by driving people over the edge.

“A successful candidate’s aim is to broaden the conversation around an issue with words that spark interest on both ends of the political spectrum… while avoiding overstating the problem,” it declared.


Both major Australian parties are failing to confront the undisputed fact that the huge costs of their climate policies CANNOT yield any beneficial result

The Australian population and economy is too small for even their most severe climate policies to have any impact on the climate

Plain-speaking Australia has been replaced by parlour games, and we all suffer. Instead of arguing big issues from first principles, politicians corral debate to avoid offending the Canberra press gallery, the ABC or theoretical swinging voters as imagined by political consultants.

This is how an election choice between orthodox economic progress and a reckless ideological punt ends up being portrayed as an ill-defined contest between evenly matched plans.

We have the unthinkable scenario of Labor refusing to reveal the economic impact of its climate and energy policies on the absurd basis that it will be less than the cost of inaction.

Our spineless national debate also means the Coalition can’t bring itself to respond with the plain fact that there is no cost to inaction; presumably because that might expose the folly of heavy costs already imposed by its climate policies. Much in this campaign is based on similar obfuscation.

Think of it this way. If Labor were to run an emotive argument on climate change during a drought — suggesting it could stop droughts, floods and bushfires — and ignoring details and cost, it just might get a head nod in focus groups. Then imagine Coalition strategists deciding they need to tackle Labor’s plan by pointing out the obvious facts about how Australia’s actions can have no environmental benefit while global emissions are rising by a much greater factor and that doubling our emissions reductions will produce serious economic disadvantages. The pollsters and consultants might step in and warn that this would be a self-defeating approach because their qualitative polling shows climate is Labor’s ground.

Instead of having a real argument about costs and benefits, the Coalition might retreat to policy and rhetoric that doesn’t confront the emotive and wrongheaded foundations of Labor’s plan but merely quibbles over the extent of action. The real debate is left ­unargued.

This, as you can see, is probably not far from what has transpired. And it is not the only debate eviscerated in such a fashion.

Too many commentators are sucked down the same path. Analysis can be constrained within the unspoken boundaries of the media/political class, unconsciously placing a greater store in resonating with peers rather than serving audiences.

This is how groupthink is formed and, together with a broad green-left bias in the media, it is why so many commentators have had to back-pedal in this campaign from earlier predictions of a Labor landslide. As we approach the crucial final week, what important facts and issues are not being aired and what myths are being perpetuated?

The economic choice is profound. At a time of global uncertainty and domestic stasis — as exemplified by the Reserve Bank’s contemplation of lowering interest rates from record emergency lows — Labor’s prescription to increase the government take from the economy by up to $387 billion across 10 years is frightening.

The major parties agree public debt levels are worrying and that returning the budget to surplus is just the start of a fiscal recovery plan. Yet the Coalition prescription is to lower taxes and constrain spending to foster growth and Labor’s plan is the polar opposite.

We know which of these works and which will lead to ruin. The postwar political and economic history of Western liberal democracies tells us the smaller government approach is required. Yet Labor promises bigger government, with more interventions in tax, wages and energy.

Labor rolled out Bob Hawke and Paul Keating to endorse its economic prescription and they rightly claimed their reforms set up decades of prosperity. But they failed to mention how their own party repudiated this aspirational approach under Kevin Rudd and is now focused not on economic ­reform but on increased taxation, redistribution and spending.

Privately, Hawke and Keating would be horrified by modern Labor’s economic regression. But the media remains conveniently ­incurious.

In economic terms this election is at least as important as the 2007 switch from John Howard and Peter Costello to Rudd and Wayne Swan. That is saying something.

The climate debate is disgraceful. There cannot be a journalist in the country who doesn’t know that global carbon emissions are growing each year by about double Australia’s annual emissions, yet they allow Shorten and Labor to get away with the fiction that their policies will be cost-free and will reduce drought, floods, bushfires and cyclones.

How is it that so many people with functioning intellects can allow this nonsense to go ­unchallenged day in, day out? Presumably they want to conform with a climate-sensitive, “woke” generation or don’t want to risk being denied access to Labor insiders on the cusp of forming government. They go with an ­orthodox fiction and fail their ­audiences.

The Greens get away with murder. Richard Di Natale has been spruiking new laws to clamp down on the media and says the aim is to stamp out hate speech, but it sounds more like silencing his ­critics. The ABC and other media have been silent about this threat to freedom of speech even though there is form — in alliance with the Greens, the Gillard government tried to impose de facto regulation on print media.

Days after these plans were exposed, Di Natale’s double standards were laid bare as he stood by candidates caught posting offensive and racist material online. One resigned but Di Natale still backs George Hanna, the Greens candidate in the Northern Territory seat of Lingiari, who referred to his indigenous Coalition opponent ­Jacinta Price as a coconut.

This is purely racist abuse — suggesting someone is brown on the outside and white on the inside — and the only excuse the Greens can offer is that Hanna, too, is indigenous. Pathetic. Imagine the reaction if the Coalition attempted to stand by a candidate in such circumstances.

Much of the media/political class buys into Labor’s paranoia about News Corp but is incurious about how the publicly funded broadcasters boost the green-left agenda 24/7 and fail to scrutinise Labor or the Greens.

They see a conspiracy when The Daily Telegraph tells the inspirational truth about Shorten’s mother but are fine with the deliberate and extended fake news around the so-called “watergate” scandal.

A clutch of 15 so-called independent MPs and candidates are boosted by the media — because they attack conservatives — but are never pinned down over who they would support to form government. They will support Labor, of course, which may prove very important as a hung parliament is a possibility.

Much of our debate seems incapable of referencing a reasonable person test. As much as the Coalition has let us down, damaged itself and made it easy for people to vote against it, it has muddled through a useful period of economic and fiscal recovery, restoring our border security and securing significant free trade deals.

Would a reasonable person believe that a switch to Labor does not involve an enormous and unnecessary risk on border security, energy affordability and reliability and economic progress?

The choice is stark and the debate is far more opaque than it should be.



For more postings from me, see  DISSECTING LEFTISM, TONGUE-TIED, EDUCATION WATCH INTERNATIONAL, POLITICAL CORRECTNESS WATCH, FOOD & HEALTH SKEPTIC and AUSTRALIAN POLITICS. Home Pages are   here or   here or   here.  Email me (John Ray) here.  

Preserving the graphics:  Most graphics on this site are hotlinked from elsewhere.  But hotlinked graphics sometimes have only a short life -- as little as a week in some cases.  After that they no longer come up.  From January 2011 on, therefore, I have posted a monthly copy of everything on this blog to a separate site where I can host text and graphics together -- which should make the graphics available even if they are no longer coming up on this site.  See  here or here


No comments: