Muzzled scientists and specious science

The motion followed closely on the heels of rallies held by the Professional Institute of the Public Service of Canada (PIPSC) across the country.  Frustrated with the government’s repudiation of the muzzling issue, PIPSC is seeking to have open and timely public access to federal scientists enshrined in their next collective agreement.

The government’s response to Dr. Hsu’s motion and PIPSC’s proposal has been stout denial. Treasury Board President Tony Clement shrugged off PIPSC’s concerns, suggesting that they are merely an attempt to “get headlines” and publicise PIPSC’s bargaining position.   

In the House, former Minister of State for Science and Technology Gary Goodyear and Minister of State for Western Economic Diversification Michelle Rempel delivered the government’s predictably dismissive response.  But less predictably, Ms. Rempel – apparently drawing on knowledge accrued “over ten years in academic research management” – elected to “use the scientific method to dissect the assertions in the motion.”

According to Ms. Rempel, the scientific hypothesis of muzzling is clearly refuted by the available evidence. Gary Goodyear agreed.  “The data do not support the Member’s motion,” said he.

So what is the evidence that has led Mr. Goodyear and Ms. Rempel to reject the muzzling hypothesis?

First, government scientists produce, on average, more than four thousand publications a year.  Second, every year science-based departments grant thousands of media interviews. Third, the current government has (apparently) spent more than $13B on funding scientific research and development.  Fourth, Canada ranks first among OECD countries in percentage of GDP allocated to so-called HERD (Higher Education Research and Development).

Therefore, conclude Mr. Goodyear and Ms. Rempel, the hypothesis is rejected: federal scientists are not being muzzled.

As proponents of the scientific method, Mr. Goodyear and Ms. Rempel should know that testing scientific hypotheses requires: (1) predictions, that is, the results one expects if the hypothesis is true; and (2) experimental results to which predictions can be compared. The hypothesis that my bathroom light doesn’t work because the bulb is burnt out leads to the prediction that if I replace it with another (working) bulb, there’ll be light.  If the hypothesis is wrong, I’ll continue to grope in the dark.

Put another way, testing a hypothesis requires that experimental results be divisible into those that are consistent with the hypothesis, and those that are not.  Instant illumination upon replacement of the old bulb is consistent with the hypothesis of a burnt-out bulb; continuing darkness is not.

In the same manner, using the number of publications or interviews, or levels of science funding, or OECD HERD performance, to test the muzzling hypothesis requires that the numbers or levels or performance that are consistent with the hypothesis be distinguished from those that are inconsistent.  No distinction, no prediction.  No prediction, no test.  No test, no evidence.

If federal scientists were being muzzled, how many publications would we expect them to produce?  How about if they’re not being muzzled? How many interviews would we expect muzzled scientists to give?  How many would we expect if they weren’t muzzled?

What level of government science funding, or what HERD ranking, is consistent with the muzzling hypothesis?  What level or ranking is inconsistent? 

Ms. Rempel and Mr. Goodyear aren’t saying.

Why not?  Because the muzzling hypothesis which forms the basis for Dr. Hsu’s motion cannot be tested with these data.  These data are, in fact, completely irrelevant.   

So not only is this contradictory evidence not contradictory, it’s not even evidence.

Yet every Conservative MP in attendance voted against Dr. Hsu’s motion.

It appears that for many of our parliamentarians, scientifically specious reasoning by the palpably uninformed trumps the scientific method every time.