Plain Packaging and "And"
Author: Michael J. McFadden
Article Published: 2011/04/14
A new study from England has found results that "may support initiatives ... that would remove branded labels to accentuate health warnings on cigarette packs."
Two particular points of interest should be noted in thinking about this piece of research. First of all, the study was conducted in a university laboratory. How the researchers were able to make the stretch to conclude anything at all about how people's microscopic eye movements might be affected in the bright and colorful real world of a convenience store or tobacco shop is beyond my understanding and probably beyond yours.
The second point of interest is that just about no matter WHAT results had been found they could have been interpreted to give results desired by antismoking advocates and grant funders.
If the researchers had found that smokers read the health warnings more frequently on packs that lacked colorful logo-names, they would have hailed it as a success showing that removing those logos and replacing them with plain packaging would make smokers more aware of health effects and more likely to quit. If they found they read them more frequently on packs WITH the colorful logo names they could have used it to argue the need for bloody pictures of corpses on the packs since the warnings were being ignored.
If they found that nonsmokers looked at the health warnings less frequently on plain packs, they could have hailed the success of educational efforts having shown that nonsmokers no longer needed to read them because they'd already absorbed the message. If they found that nonsmokers looked at health warnings on packs more frequently if they weren't distracted by looking at brand logos they could argue that the plain packs were more effective in preventing them from suddenly deciding to buy a pack and start smoking.
The category "light smokers" was also included in the research but neither the article nor the abstract separated them out sufficiently to be able to say anything about them. Instead they presented the research as smokers vs. "nonsmokers AND light smokers" which rang a very suspicious bell in my mind: it's the same sort of phrasing that got Dr. Elizabeth Klein's original study on bar employment and smoking bans headlines around the world. The Klein study lumped both bar AND restaurant employees together in order to conclude that "Bans do not hurt bar AND restaurant employment." A more careful analysis of the data, indeed even a fairly superficial analysis of the data that the researchers actually had in front of them in raw form, showed pretty clearly that the word "and" was being used to cover up the damage to bar employment so that bars in proposed ban areas wouldn't fight the bans. See the commentary after Jacob Grier's excellent column for the full story on that and for a true appreciation of how Antismokers can play with such a simple word as "and" in order to convey an impression that can be very different from the actual research findings.
So, overall, what did this study show? Without having bought the study for full analysis, I would say that the most likely thing it showed was that researchers can design a study and take whatever data they want to from it to support whatever happens to be the desired political goal.
The full Reuters article on the study and link to its abstract is available here. You can view the full study itself if you or your funders are willing to spend $35 for the privilege of viewing it for a 24 hour period. Just fill out the necessary forms here.
By the way, the results of the study, no surprise, showed that replacing normal packaging with "plain packaging" would be a "good" thing. While there seemed to be no effect on smokers, the nonsmokers moved from a 14/14 glance split to a 16/12 glance split ... i.e. two extra glances at the health warnings out of 28 total glances. The figures for "light smokers" alone without the "and" connection were simply not revealed.