Setting the record straight on some issues.

Smoking will kill you, even if someone else is doing the smoking, right? No. Call it what you will - passive smoking, second hand smoke, whatever - it isn't dangerous. Think about it. Smokers are drawing the smoke directly into their lungs. Anything that others get has been diffusing into the air. For more information, see here.

Even if the results are accepted at face value, the impressive-sounding risk figures for lung cancer and heart disease imply that passive smoking accounts annually for one extra death in every 10,000.

This article puts the whole thing in perspective.

Habitual, lifelong smokers face a 30 to 40-fold higher risk of contracting lung cancer than non-smokers. That sounds massive and many smokers are persuaded to quit because they believe it is. But, since the risk of lung cancer in non-smokers is minuscule, it does not amount to an objectively high risk.

Sandford admits: "Smokers are more likely to die of heart disease than lung cancer." And pro-smoking campaigner Joe Jackson argues: "Even if you're a heavy smoker, your chances of not getting lung cancer are still more than 99 percent.

And my favorite quote:

Dr Ken Denson of the Thame Thrombosis and Haemostasis Research Foundation says: "I simply do not know where they conjure up their statistics. The statistics for passive smoking in particular would not be published or even considered in any other scientific discipline.

BTW, Shaken Baby Syndrome doesn't exist, either. A new study found that an adult man or woman cannot shake a baby hard enough to cause internal damage. If a baby is shaken it will damage the neck, not the brain.

And the arctic is not warming.

"Antarctica has been cooling for the last 50 years. Most of the Arctic has not warmed over long time scales," she told the news service. "Temperatures (have) always changed in the past and (they) always will. . . . We don't have enough understanding of natural variability and we don't see enormous amounts of temperature change to be alarmed about," she explained.

And this:

Naurzbaev, et al (2002) created a proxy temperature data set spanning nearly 2,500 years for the Taimyr Peninsula of northern Russia, all of which is poleward of 70° N. The authors studied tree rings-widths of living and deceased larch trees. They reported that "the warmest periods over the last two millennia in this region were clearly in the third, tenth to twelfth and during the twentieth centuries." The first two, they claim, were warmer than those of the last century. Twentieth century temperatures appeared to peak around 1940.

...

The ACIA appears to be guilty of selective use of data. Many of the trends described in the document begin in the 1960s or 1970s -- cool decades in much of the world -- and end in the warmer 1990s or early 2000s. So, for example, temperatures have warmed in the last 40 years, and the implication, "if present trends continue," is that massive warming will occur in the next century. Yet data are readily available for the 1930s and early 1940s, when temperatures were comparable to (and probably higher than) those observed today. Why not start the trend there? Because there is no net warming over the last 65 years?

For that matter, can't we dispense with the use of linear trends for cyclical time series which have a cyclical nature? My college statistics prof would have been very upset at this practice, because the character of a trend line in a data set like the one shown in Figure 3 is largely a function of the starting and ending points selected.

I also looked closely at many of the charts and saw misleading information. For example, the chart on global sea level rise goes back only 10 years but shows a steep increase. Then I read the y-axis -- a total rise of about one inch! Since we don't know where the data originated (the caption says "from a satellite launched in 1992") we can only wonder whether the measurement accuracy is sufficient to even measure a one inch change (or whether such a change even matters!).