M Goldberg, et al. (1995)

When Packages Can't Speak: Possible Impacts of Plain and Generic Packaging of Tobacco Products was a 1995 report based on a series of surveys, for Health Canada, of teenagers, over 5 different studies, to examine the potential effect plain packaging might have on

• the uptake of smoking to begin with,

• the impact on the recognition of, and the ability to remember, the warnings on packaging,
• the probability of stopping smoking

For the Direct Questioning surveys, Padilla & Watson(2008) summarised that

• Teenagers have mixed views on what they believe to be the impact of generic [plain] packaging.
• Results suggest that the effects of generic [plain] packaging on smoking would be marginal.

In the Visual Image survey that

• Generic packaging increased the recall rate of only one of three health warnings. The authors suggest that the exposure time was too short and that these results cannot be extrapolated to a more natural long term-setting.

And in the Conjoint survey that

• Packaging is generally as important as brand influence and peer influences (except for teenage non-smokers).
• Results suggest that plain and generic packaging will, to some “unknown degree, encourage non-smokers not to start smoking and smokers to stop smoking”

National Survey

The purpose of the National Survey was to assess the knowledge, attitudes and beliefs held by teenagers (14-17 years) regarding smoking, brands, brand images, generic packaging and perceived impact of such packaging on teenagers. From an initial pool of 6,213 teenagers, 1,200 were eventually questioned from 14 Canadian cities. Those not eventually questioned dropped out because they didn't fit the age criteria, they were overt anti-smokers, or they simply refused to answer the questionairre.

Padilla & Watson(2008) concluded that

The objective of the survey was simply to provide descriptive statistics on smoking beliefs, patterns and behaviours of teens in Canada. No attempt was made to establish a causal relationship between cigarette packaging and youth smoking.[Page 40]

Word Image Survey

The purpose of this survey was to determine what, if any, differences were perceived between branded packets, plain packets or packets with a "lungs" symbol. The cohort for this study seems to have been the same cohort as the National Survey, since 1,200 took part, and they answered the same screening questions.

Padilla & Watson(2008) concluded that

The results of this survey cannot be used to inform a policy decision regarding implementation of generic packaging. The likely effects of generic packaging can only be inferred from a comparison of the situation before and after the measure.[Page 42]

Visual Image Survey

Pictures of 6 (types of) people were shown with different package types (3 brands, and pack types of branded, plain or plain+"lungs",) in the lower right hand corner. Participants were then asked to agree/disagree on a 5 point scale with the statement “Consider this (picture). Is (brand name) in this package right or wrong for this (woman/man)”. The brands selected were from a previous survey for which teenagers had the greatest convergant images.

Padilla & Watson(2008) again:

By selecting only those brands that were more strongly associated with a certain person-type in the national survey, the results of this analysis may have overestimated the actual link between brands and perceived images.[Page 43]

Recall and Recognition Survey

This survey was of 400 Vancouver teenage smokers, to assess what differences plain/branded packs had on how much attention was paid to both the brand, and any warning messages present on the packs.

3 images of a table with a packet of cigarrettes, different brand each time, (with the pack either always branded or always plain), a magazine, can of pop and a bottle of headache pills. Images were shown for 4 seconds, and the respondants asked to list what they could see, the brand of cigarettes, and the warning on the packet.

The respondants were then shown all three packs with the warnings hidden, and asked to name which warning went on which pack.

Padilla & Watson(2008):

The results of the study do not indicate that health warnings are better recalled when displayed on generic packages.

Conjoint survey

Respondants, which was the same cohort that participated in the Recall and Recognition survey above, were asked to choose between alternatives that differed in

• package type,

• brand,

• price, and

• peer influence (friends smoke/do not smoke).

Goldberg et al (1995) themselves concluded the extent of the influence of generic packaging on smoking decisions:

cannot be validly determined by research that is dependent on asking questions about what they think or what they might do if all cigarettes sold in the same generic packages.[Page 129]