Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2017 Jul 1.
Published in final edited form as: Drug Alcohol Depend. 2016 May 2;164:97–105. doi: 10.1016/j.drugalcdep.2016.04.031

Using Eye-tracking to Examine How Embedding Risk Corrective Statements Improves Cigarette Risk Beliefs: Implications for Tobacco Regulatory Policy

Kirsten Lochbuehler a,b,*, Kathy Z Tang d, Valentina Souprountchouk d, Dana Campetti d, Joseph N Cappella a,b, Lynn T Kozlowski c, Andrew A Strasser a,d
PMCID: PMC4893893  NIHMSID: NIHMS785096  PMID: 27160034

Abstract

Background

Tobacco companies have deliberately used explicit and implicit misleading information in marketing campaigns. The aim of the current study was to experimentally investigate whether the editing of explicit and implicit content of a print advertisement improves smokers’ risk beliefs and smokers’ knowledge of explicit and implicit information.

Methods

Using a 2(explicit/implicit) x 2(accurate/misleading) between-subject design, 203 smokers were randomly assigned to one of four advertisement conditions. The manipulation of graphic content was examined as an implicit factor to convey product harm. The inclusion of a text corrective in the body of the ad was defined as the manipulated explicit factor. Participants’ eye movements and risk beliefs/recall were measured during and after ad exposure, respectively.

Results

Results indicate that exposure to a text corrective decreases false beliefs about the product (p < .01) and improves correct recall of information provided by the corrective (p < .05). Accurate graphic content did not alter the harmfulness of the product. Independent of condition, smokers who focused longer on the warning label made fewer false inferences about the product (p = .01) and were more likely to correctly recall the warning information (p < .01). Nonetheless, most smokers largely ignored the text warning.

Conclusions

Embedding a corrective statement in the body of the ad is an effective strategy to convey health information to consumers, which can be mandated under the Tobacco Control Act (2009). Eye-tracking results objectively demonstrate that text-only warnings are not viewed by smokers, thus minimizing their effectiveness for conveying risk information.

Keywords: Eye-tracking, Tobacco advertising, Tobacco regulatory policy

1. INTRODUCTION

An increased public awareness of health risks of smoking has led the tobacco industry to develop “new” products that have been marketed as less harmful than regular cigarettes. The development, advertising and marketing of “new” products started with the promotion of light cigarettes in the 1960s (National Cancer Institute, 2001), and continued with the marketing of potentially reduced exposure products (PREPs), to the most recent marketing of Modified Risk Tobacco Products (MRTPs), which are often developed as lower tar/nicotine cigarettes (Dunsby and Bero, 2004; US Department of Health Human Services, 2000). In the aggressive marketing of harm reduction, tobacco companies have effectively used cigarette pack design, colors, labels and descriptive terms to communicate strength, harshness, lower nicotine, tar levels and risk of their products (Anderson et al., 2006; Bansal-Travers et al., 2011; Philip Morris Co; Philip Morris Co; Pollay and Dewhirst, 2001; 2002; Slade, 1997; Wakefield et al., 2002).

While previous efforts under the Tobacco Control Act (TCA; 2009) have successfully mandated removal of descriptive terms, other actions to convey health information to consumers, for example the implementation of graphic warning labels, have been upheld by U.S. federal courts. Given evidence that many smokers misinterpret the information contained in cigarette marketing and perceive a modified type of cigarette as safer or having lower health risks than regular cigarettes (Hamilton et al., 2004; O’Connor et al., 2005; Parascandola et al., 2009; Shadel et al., 2006; Shiffman et al., 2007, 2004), it is important to develop more accurate advertising and to counter misinformation in order to better inform smokers of their risks. This correction of misinformation in advertisements calls for and requires a better understanding of how accurate information can be conveyed to smokers. The TCA empowered the Food and Drug Administration (FDA) to review MRTPs to ensure that marketing and advertising claims are supported by scientific evidence and are not misleading (U.S. FDA Center for Tobacco Products, 2015). Further, in 2006, the U.S. District Court ordered the use of corrective statements in advertising and promotion to correct past deceptive practices of the tobacco companies and to enhance knowledge among consumers (Smith et al., 2011; U.S. District Court DC. U.S. v. Philip Morris USA, 2006).

Few studies have investigated the effects of counter-advertising and corrective messages on risk perception of cigarette products (Biener et al., 2007; Kozlowski et al., 2001, 1999, 2000; Shiffman et al., 2001; Strasser et al., 2008; Tangari et al., 2010). Some studies demonstrated that counter-marketing ads on TV (Kozlowski et al., 2000) and radio (Kozlowski et al., 2001, 1999) can improve the risk perception of light cigarettes, and that corrective print advertisements can influence consumers’ beliefs about smoking (Tangari et al., 2010). With regard to correctives in print advertisements, Biener et al. (2007) found that providing health information (adding explicit content) in the form of a box at the bottom of print advertisements had only little impact on the risk perception of the product. It remains unknown whether and how much attention smokers paid to the health information that was provided in a box similar to a text warning label at the bottom of the ad. Previous research has shown that viewers rarely focus on a text warning label (Fischer et al., 1989; Munafò et al., 2011). Given these results and the fact that the center of the ad is viewed more attentively (Strasser et al., 2012), integrating explicit corrective text messages in the body of the ad might be a more effective strategy. In a study where smokers were either exposed to the original Quest print advertisement, an ad in which the cigarette packages were altered to the color red (altering implicit content), or an ad where all text was removed from the ad (removing explicit content), it was found that removing the text affects beliefs about the harmfulness of the product (Strasser et al., 2008). Smokers who were exposed to the altered implicit ad did not report fewer false beliefs about the harmfulness of the product than smokers exposed to the original advertisement. It is plausible but remains unknown whether an ad with combined explicit and implicit accurate content would be associated with a more accurate understanding of harm exposure.

These studies demonstrate that the risk perception of a product can be corrected by altering explicit and implicit content of the ad. However, it remains unclear whether adding explicit accurate information (i.e., a corrective statement about product risk) in the body of the ad and altering implicit information (i.e., cigarette package color) affects smokers’ risk perception of the product. Therefore, using a 2(explicit/implicit) x 2(accurate/misleading) between-subject design, we investigated the effect of manipulating advertisement content on smokers’ false beliefs about the product and smokers’ recall of explicit and implicit information after viewing the advertisements. The manipulated explicit factor was the inclusion of a text corrective about tar and nicotine content which has been shown to have a sustained effect on perceived product risk (Kozlowski and Sweeney, 1997). The manipulation of the cigarette package color and background images that have been shown to impact risk perception and have been increasingly used to imply health claims (Bansal-Travers et al., 2011) were examined as an implicit factor.

Additionally, to better understand how smokers view implicit and explicit content, we assessed attention to different parts of the advertisement using eye-tracking. We hypothesized that smokers who were exposed to the explicit and implicit accurate ads, compared to those exposed to the misleading ads, will report fewer false beliefs about the product and will better recall the information stated by the corrective. The results of the present study will provide valuable information on what content of the advertisement draws attention, what content distract smokers from viewing health information, where corrective health information needs to be placed in an ad, and how implicit and explicit information in cigarette advertisements affect risk beliefs about nicotine products. Given the continuous introduction of nicotine-containing products (e.g., low nicotine cigarettes, little cigars, e-cigarettes) and the significant marketing of these products, it is critical that regulatory efforts are aware of how product users can be mislead.

2. MATERIAL AND METHODS

2.1 Sample and procedure

A total of 203 current smokers participated in the study. Inclusion criteria were: currently smoking a minimum of 10 cigarettes/day (non-menthol) for a minimum of 5 years; not currently trying to quit or intending to do so in the next two months; between 21 - 65 years old; speaking English fluently; no current substance abuse; no visual impairments. Participants responding to advertisements were screened for eligibility by telephone and those deemed eligible were scheduled. The study consisted of a 90-minute single session where participants were randomly assigned to view one of four versions of the Quest advertisement while eye movements (EM) were measured. After a research assistant explained the procedure and demonstrated the equipment, participants gave informed consent and completed baseline questionnaires (see demographic and smoking history measures). Then, participants provided a carbon monoxide breath sample to biochemically verify smoking status. Based on sensitivity analysis by Raiff et al. (2010), a 5 parts per million (ppm) threshold was used for study inclusion. None of the participants fell under this threshold. Then, participants smoked one of their own cigarettes to standardize time since their last cigarette. They were seated in front of the eye-tracking device, were calibrated and asked to view the advertisement for 30 seconds. After viewing, participants completed post-advertisement measures (risk belief and recall measures). Finally, participants were debriefed to the purpose of the study, had any questions answered and received $50 compensation. The protocol was approved by the University of Pennsylvania Institutional Review Board.

2.2 Advertisement design

We used a 2(explicit/implicit) x 2(accurate/misleading) between-subject design (see Figure 1 for an illustration of the advertisement conditions). The inclusion of a corrective statement in the body of the ad was investigated as an explicit factor to decrease misinformation about the product. The explicit misleading conditions portrayed the text of the current Quest advertisement, which has been associated with smokers making false inferences about the product (Shadel et al., 2006; Strasser et al., 2008). The explicit accurate conditions were altered and the corrective statement “Nicotine free does not mean risk free. Quest contains as much tar as a light cigarette” replaced the text “Real cigarettes, real premium tobacco, real smoking enjoyment” in the body of the advertisement. The theme of the corrective statement was chosen because this message had been found effective, recalled easily in previous research (Kozlowski et al., 1999), and provided information about nicotine, risk and tar which are central themes to cigarette harm. The rest of the text remained the same as in the current Quest advertisement.

Figure 1.

Figure 1

Open in a new tab

Advertisement conditions.

The manipulation of graphic features was examined as an implicit factor to convey product harm. The implicit misleading conditions used the graphics of the current version of the Quest advertisement (image D), which shows blue colored cigarette packs diagonally on the “road to nicotine-free”. The packaging color and background imply light and safe (Bansal-Travers et al., 2011), sun and clouds constitute the background of the ad. The advertisement used in the implicit accurate conditions was altered such that cigarette packages were presented in red color, laterally across the advertisement. The background clouds and sun were removed, implying more accurate harmfulness of the product, in contrast to implied lightness or less harm.

2.3 Measures

2.3.1 Eye-tracking

Gazetracker software (v.07.01.243.128) was used to display the advertisements. EM were assessed by an Eye-Trac 6 control unit with an R6 pan/tilt optics system and video head tracker (Applied Science Laboratories, Boston, MA). Data transfer was linked using the Eye-Trac 6 User Interface Program (v.1.30.8.0). Areas of interest (AOIs) were identifıed a priori (see Figure 2 for an illustration of the AOIs). For each region, dwell time (total time viewed in AOI; seconds) and time to fırst viewing of AOI (seconds) were assessed (Strasser et al., 2012). Fixations were operationalized as any 60-pixel-diameter space with three consecutively sampled observations with a minimum cumulative 200ms duration, consistent with previous viewing images and reading tasks (Rayner, 1978).

Figure 2.

Figure 2

Open in a new tab

Areas of interest.

2.3.2 Risk Beliefs

To assess risk beliefs, we used an 8-item, 5-point scale (1 = definitely untrue; 5 = definitely true) (Shadel et al., 2006; Strasser et al., 2008). The 8 items included: Quest cigarettes: a) are lower in nicotine than regular cigarettes (true item); b) are lower in tar than regular cigarettes; c) are less addictive than regular cigarettes; d) are less likely to cause cancer than regular cigarettes; e) have fewer chemicals than regular cigarettes; f) are healthier than regular cigarettes; g) make smoking safer; and h) help people quit smoking. Items b-h were considered false statements, based on information in the advertisements indicating that tar levels are not reduced and that Quest cigarettes are not intended to assist smokers with quitting. A belief scale was constructed on items b-h where higher scores mean making more false inferences about Quest cigarettes (α = .72).

2.3.3 Recall

Five items in multiple choice format were used to assess the recognition of central explicit and implicit messages conveyed in the advertisement viewed: 1) Quest cigarettes contain no nicotine at all; 2) What statement best captures the content of the warning label?; 3) What color were the cigarette packs?; 4) How did the cigarette packs appear in the ad?; 5) Quest cigarettes contain as much tar as a light cigarette. Responses to each recall item were scored as correct or incorrect.

2.3.4 Demographic and Smoking History Measures

Age, gender, race, ethnicity, educational background, years smoking, daily cigarette consumption, age of first cigarette, time since last cigarette prior to the session, number of cigarettes smoked on session day, and nicotine dependence (Heatherton et al., 1991) were assessed.

2.4 Preparation of eye movement data

There were 26 sessions (12%) with low quality EM data that were excluded from all analyses, resulting in a final sample of 177 smokers. Exclusion was due to calibration difficulties, technical problems with the eye-tracking equipment, or because fixations on the ad were recorded less than 66.6% of the time. The excluded participants were equally distributed over conditions and did not differ by smoking or descriptive measures from those participants included in the analyses. Viewing time of the whole advertisement was examined to verify attention to the screen during presentation and did not differ by condition.

2.5 Analysis plan

Data were analyzed in SPSS. Descriptive statistics were used to characterize the overall sample and each advertisement group. Analysis of variance (ANOVAs) and Chi-square tests (χ2) were conducted to identify potential group differences. Analyses of covariance (ANCOVAs) were used to examine the effect of advertisement type on attention to the areas of interest (AOIs) “corrective statement,” “cigarette packs,” and “warning label.” ANCOVAs and logistic regression analyses were conducted to determine models of predicting false inferences about the product and correct recall of information of the corrective and the warning label. Attention measures were inserted into each model. In each analysis we controlled for age, gender, daily cigarette consumption, and the time of the last cigarette smoked prior to the session.

3. RESULTS

3.1 Descriptive statistics and randomization verification

Table 1 displays sample characteristics by condition. Participants were on average 30.6 years old (SD = 8.07; range = 21 - 58). The sample was predominantly male (64.5%) and Caucasian (92%). Most participants had completed a high school degree (98%); 39.4% had completed college. Participants on average reported smoking for 12.86 years (SD = 6.8), and smoking 16.6 cigarettes/day (SD = 5.7), and an average nicotine dependence score of 5.1 (SD = 2.1). The most popular brands were Marlboro (41.7%), Camel (24.6%), Parliament (15.4%), and American Spirit (8%). CO-levels measured at session commencement ranged between 6 - 54 ppm (M = 19.32; SD = 9.36). ANOVAs and χ2-tests were conducted to examine whether randomization over the four conditions was successful. Analyses only showed differences with regard to daily cigarette consumption.

Table 1.

Descriptive statistics and smoking characteristics by condition.

Characteristics Explicit
accurate -
Implicit
accurate		 Explicit
accurate -
Implicit
misleading		 Explicit
misleading -
Implicit
accurate		 Explicit
misleading -
Implicit
misleading
N 47 44 42 44
Gender male (count) 30 27 26 28
Age (M, SD) 31.36 (9.19) 28.30 (6.92) 32.33 (8.30) 29.18 (6.69)
Daily cigarettes (M, SD) 18.32 (5.87) 16.75 (6.24) 15.31 (4.87) 15.32 (5.32)
Years smoking (M, SD) 13.57 (7.55) 11.30 (5.66) 13.76 (6.94) 12.07 (6.29)
FTND (M, SD) 5.60 (2.07) 5.07 (2.15) 4.86 (2.11) 4.68 (1.95)
CO-level (M, SD) 20.30 (7.92) 22.05 (10.92) 21.43 (11.95) 18.98 (9.55)
Time since last cigarette
prior to session (M, SD)		 65.55 (118.46) 42.39 (19.46) 39.57 (26.77) 55.73 (103.95)
Age of first cigarette (M, SD) 17.13 (3.03) 15.84 (2.77) 17.38 (5.33) 15.91 (2.14)
Number of cigarettes smoked
on session day (M, SD)		 4.43 (2.76) 4.09 (2.46) 3.90 (2.27) 3.77 (2.61)
Race (count)
 Caucasian		 42 40 35 40
 Asian
 African-American
 More than one race		 1
2
0		 1
0
0		 2
2
1		 2
1
1
Education (count)
 High School
  College
39
15
36
18
36
23
38
17
Open in a new tab

Note. Boldface indicates statistical significance (p < .05).

3.2 Effects on attention

ANCOVAs were conducted to examine the effect of advertisement type on latency to and dwell time of the AOIs ”corrective statement,” ”cigarette packs,” and ”warning label.” Table 2 displays the dwell time of and the latency to AOIs overall and by advertisement type. Attention data showed that smokers focused on the AOI ”corrective statement” relatively early in the overall viewing pattern (M = 2.16; SD = 3.32), while the latency to the AOI “warning label” was relatively long (M = 9.27; SD = 7.40). The results also showed an effect of explicit condition on the dwell time of and the latency to the AOI ”corrective statement.” Smokers in the explicit accurate conditions spent more time focusing on the AOI ”corrective statement” than smokers in the explicit misleading conditions (M = .93s vs. M = .57s; F (1,169) = 7.80; p = .006, η2 = .04). Smokers in the implicit accurate conditions fixated on the AOI ”corrective statement” faster than smokers in the implicit misleading conditions (M = 1.53s vs. M = 2.79s; F (1,168) = 8.19; p = .005, η2 = .046). No effect of condition on the latency or dwell time of the AOI ”warning label” or ”cigarette packs” was found.

Table 2.

Eye-tracking measures overall and by advertisement type.

Area of interest and
measures		 Explicit
accurate
conditions
M (SD)		 Explicit
misleading
conditions
M (SD)		 Implicit
accurate
conditions
M (SD)		 Implicit
misleading
conditions
M (SD)		 Overall/
Independent
of condition
M (SD)
Cig packs (4)
Dwell time (in s) 8.31 (3.22) 8.52 (3.89) 8.79 (3.73) 8.03 (3.33) 8.41 (3.55)
Latency .59 (1.64) .45 (1.21) .47 (1.39) .57 (1.51) .52 (1.44)
Warning labels (6)
Dwell time 1.51 (1.69) 1.95 (2.00) 1.50 (1.79) 1.95 (1.90) 1.72 (1.85)
Latency 9.79 (7.81) 8.71 (6.94) 10.49 (7.26) 8.05 (7.39) 9.27 (7.40)
Corrective statement (7)
Dwell time .93 (.93) .57 (.70) .77 (.89) .74 (.79) .75 (.84)
Latency 2.34 (3.27) 1.97 (3.38) 1.53 (1.90) 2.79 (4.21) 2.16 (3.32)
Open in a new tab

Note. Attention measures in seconds.

Boldface indicates statistical significance (p < .01).

3.3 Effects on risk beliefs

An ANCOVA was performed to investigate the effect of advertisement type on risk beliefs (see Table 3 and 4). There was a main effect of the explicit condition (F (1,169) = 7.26, p = .008, η2 = .04), such that smokers who were exposed to explicit misleading ads (M = 3.01; SD = .64) reported more false beliefs than those exposed to explicit accurate ads (M = 2.79; SD = .62). Adding dwell time of or latency to the AOI ”corrective statement” to the model, the results again showed a main effect of the explicit condition. Dwell time of and latency to the AOI ”corrective statement” did not affect risk beliefs. An ANCOVA with dwell time to the AOI ”warning label” and condition as independent variables and risk beliefs as dependent variable indicated a main effect of dwell time to the AOI ”warning label” (F (1,165) = 6.25, p = .01, η2 = .04) and a trend of explicit condition (F (1,165) = 3.47, p = .06, η2 = .02). Smokers who spent more time looking at the AOI ”warning label” and smokers who were exposed to the explicit accurate conditions made fewer false inferences about the product.

Table 3.

Analysis of Covariance for risk beliefs.

Source Sum of
Squares		 df Mean
Square		 F p η 2
Explicit 138.36 1 138.36 7.26 .008 .041
Implicit 6.542 1 6.542 .343 .56 .002
Explicit*implicit 2.762 1 2.76 .145 .70 .001
Error 3221.26 169 19.06
Total 76361.00 177
Explicit 1.745 1 1.745 4.38 .04 .03
Implicit .123 1 .123 .309 .58 .002
Explicit*implicit .010 1 .010 .026 .87 .000
Dwell corrective .023 1 .023 .057 .81 .000
Error 65.689 165 .398
Total 1558.39 177
Explicit 1.952 1 1.952 4.916 .03 .029
Implicit .003 1 .003 .009 .93 .000
Explicit*implicit .106 1 .106 .266 .607 .002
Latency corrective .568 1 .568 1.431 .233 .009
Error 65.105 164 .397
Total 1548.51 176
Explicit 65.173 1 65.173 3.469 .06 .021
Implicit .487 1 .487 .026 .87 .000
Explicit*implicit .026 1 .026 .001 .97 .000
Dwell warning label 117.363 1 117.363 6.247 .01 .036
Error 3100.02 165 18.788
Total 76361.0 177
Open in a new tab

Note. Adjusted for age, gender, daily cigarette consumption, and the time of the last cigarette smoked prior to the session.

Table 4.

Analysis of Covariance for individual risk beliefs.

Risk belief
Quest
cigarettes:		 Source Sum of Squares df Mean
Square		 F p η 2
are lower in
nicotine
than regular
cigarettes		 Explicit .152 1 .152 .18 .67 .001
Implicit .160 1 .160 .19 .66 .001
Explicit*implicit .497 1 .497 .60 .44 .004
Error 140.02 169 .83
Total 3170.00 177
are lower in
tar than
regular
cigarettes		 Explicit 5.190 1 5.190 4.40 .04 .025
Implicit .51 1 .51 .430 .51 .003
Explicit*implicit .065 1 .065 .055 .81 .000
Error 199.272 169 .398
Total 2027.00 177
are less
addictive
than regular
cigarettes		 Explicit 3.125 1 3.125 3.135 .08 .018
Implicit .009 1 .009 .009 .75 .001
Explicit*implicit .162 1 .162 .162 .688 .001
Error 168.455 169 .997
Total 1688.00 177
are less
likely to
cause cancer
than regular
cigarettes		 Explicit 2.209 1 2.209 2.447 .12 .014
Implicit .004 1 .004 .004 .95 .000
Explicit*implicit .622 1 .622 .690 .41 .004
Error 152.553 169 .903
Total 1133.00 177
have fewer
chemicals
than regular
bcigarettes		 Explicit 2.310 1 2.310 2.972 .09 .017
Implicit .486 1 .486 .625 .43 .004
Explicit*implicit .000 1 .000 .000 .98 .000
Error 131.361 169 .777
Total 2045.00 177
are healthier
than regular
cigarettes		 Explicit 2.702 1 2.702 2.282 .13 .013
Implicit .027 1 .027 .023 .88 .000
Explicit*implicit .021 1 .021 .017 .90 .000
Error 200.158 169 1.184
Total 210.305 177
make
smoking
safer		 Explicit 4.076 1 4.076 3.604 .06 .021
Implicit .670 1 .670 .593 .44 .003
Explicit*implicit .537 1 .537 .475 .49 .003
Error 191.113 169 1.131
Total 1336.00 177
help people
quit
smoking		 Explicit 1.099 1 1.099 .794 .37 .005
Implicit .014 1 .014 .010 .92 .000
Explicit*implicit .016 1 .016 .011 .92 .000
Error 233.870 169 1.384
Total 2170.00 177
Open in a new tab

Note. Adjusted for age, gender, daily cigarette consumption, and the time of the last cigarette smoked prior to the session.

3.4 Effects on recall of information of the corrective and the warning label

In total, 87% of participants correctly recalled item 1 (nicotine content), 42.9% correctly recalled item 2 (warning label content), 93.8% correctly recalled item 3 (pack color), 57.6% correctly recalled item 4 (pack appearance), and 61.6% correctly recalled item 5 (tar content). No differences between conditions with regard to correct recall of item 1, 2, 3, and 4 were found. Table 5 displays the results of logistic regression models assessing the association between condition, attention and recall of the information of the corrective (tar content) as well as the warning label. Explicit condition significantly predicted correct recall of the tar content (OR = 2.42, 95% CI = 1.00, 5.87; p = .05) in a model that also included implicit condition and dwell time to the AOI “corrective statement.” Smokers who were exposed to the explicit accurate conditions were more likely to correctly recall the tar content than smokers who were exposed to the explicit misleading conditions. In a similar model including latency to the AOI “corrective statement” explicit condition (OR = 2.70, 95% CI = 1.18, 6.18; p = .02) also significantly predicted correct recall. Dwell time of and latency to the AOI “corrective statement” were unrelated to correct recall of the tar content. Correct recall of the warning label content was significantly associated with dwell time of the AOI “warning label” (OR = 2.26, 95% CI = 1.45, 3.53; p = .00). Smokers who focused longer on the AOI ”warning label” were more likely to correctly recall the content of the warning label. In a similar model, latency to the AOI “warning label” (OR = 1.00, 95% CI = .92, 1.09; p = .98) did not predict correct recall of the content of the warning label.

Table 5.

Logistic regression analysis predicting recall of tar content and warning label content.

Recall tar content
OR CI p
Explicit 2.42 1.00, 5.87 .05
Implicit .76 .31, 1.84 .54
Dwell corrective .87 .33, 2.30 .77
Explicit 2.70 1.18, 6.18 .02
Implicit 1.07 .48, 2.39 .88
Latency corrective 1.18 .95, 1.45 .13
Recall warning label content
OR CI p
Explicit 1.98 .73, 5.33 .18
Implicit 1.60 .58, 4.38 .36
Dwell warning label 2.26 1.45, 3.53 .00
Explicit 1.66 .58, 4.72 .35
Implicit .59 .21, 1.68 .32
Latency warning label 1.00 .92, 1.09 .98
Open in a new tab

Note. Adjusted for age, gender, daily cigarette consumption, and the time of the last cigarette smoked prior to the session.

4. DISCUSSION

This study examined whether the correction of explicit and implicit content of an advertisement affects attention to the ad, false beliefs about the product, and recall of information provided by the ad. Results indicate that smokers focus relatively early in their viewing pattern on an explicit text corrective that was embedded in the body of the ad. In addition, exposure to the corrective statement improves smokers’ beliefs about the harmfulness of the product and recall of the information provided by the corrective. Interestingly, attention duration to the corrective did not affect these findings, suggesting that smokers easily acquire this information. These results provide novel, objective information on where to optimally place health information in print ads.

Smokers who focused longer on the warning label made fewer false inferences about the product and were more likely to correctly recall the warning label information. However, most smokers gave little attention to the warning label, and the latency to the warning label was relatively long, exceeding the average total viewing time of a print ad in a real-life situation (Lohse, 1997). In addition, only 41% of participants were able to correctly recall the statement of the warning label. Moreover, attention to the warning label was not improved by either explicit or implicit manipulations of the ad. Together, this pattern of results indicates that while the content of the text-only warning is information smokers are capable of recalling, the late focus on the warning labels makes them highly ineffective as a way to convey information. Given that smokers rarely view text-only warning labels (Fischer et al., 1989; Strasser et al., 2012), the results of this study recommend to embed health information in the center of an ad, an area that is viewed more attentively or to consider warnings with graphic content that can affect attention.

In this study the accurate altered graphic features unsuccessfully conveyed the harm of the product. A single exposure was insufficient to affect beliefs, but repeated exposure could reveal a different result. However, smokers focused faster on the corrective in ads that were more graphically accurate. Possibly, the misleading graphic content distracts and prevents viewers from focusing faster on other parts of the ad. A shorter latency to the corrective might be particularly important in real-life situations, in which viewers generally spend significantly less time on a print ad than in the current study (Lohse, 1997).

These results also inform the marketing of new tobacco and other products containing nicotine. Conveying less harm has been a recurring theme (Pederson and Nelson, 2007) and it will likely be used again by the tobacco industry when new products enter the marketplace. There is already some evidence that similar efforts have recently been made in the marketing of e-cigarettes (de Andrade et al., 2013; Grana and Ling, 2014; Huang et al., 2014; Luo et al., 2014; Paek et al., 2014; Richardson et al., 2013, 2014). Reduced-nicotine cigarettes might become another product that will be targeted and marketed by the tobacco industry as less harmful than regular cigarettes. Recent research has shown that reduced-nicotine cigarettes decreased nicotine exposure and dependence and the number of cigarettes smoked among smokers (Donny et al., 2015). In an environment where nicotine content is regulated, it can be expected that the tobacco industry will aggressively market these products to increase their appeal. Research by Donny et al. (2015) used investigational cigarettes that carried no labeling or package information and those cigarettes were not supported by advertising or marketing, which is a significant focus of the tobacco industry. The TCA (2009) allows the FDA to regulate nicotine levels in cigarettes as well as the way tobacco products are marketed and labeled. The results of the current study provide highly relevant, objective data regarding the correction of misinformation of advertisements of current tobacco products. They also inform the emerging public health question of the marketing of new tobacco products.

A number of limitations to this study should be acknowledged. First, in the current study, the warning label on the print ads were text-based only. It remains unknown whether a corrective has the same effects in the presence of a graphic warning label which has been shown to be attention-grabbing (Strasser et al., 2012). Future research should investigate whether the combination of a corrective and a graphic warning label increases the effects or whether they compete for viewers’ attention. Second, a single exposure to an ad with more accurate graphic features (i.e., cigarette package color) had no effect on risk beliefs or recall. However, this is not to conclude that any implicit (graphic) manipulation would fail in conveying the harmfulness of the product. Further research of the effects of other types of graphic manipulations on product perception is needed. Third, the current sample consisted of predominantly well-educated, male, Caucasian smokers, and different effects may be found if we were to focus on different groups of smokers. Further studies on how a potential user might engage with an ad for a new product including different groups of smokers would be of great value and are therefore suggested.

The findings of the current study provide important implications for mandating corrective information in print ads, which falls under the regulatory authority of the FDA. Embedding a corrective statement in the body of an ad has small but reliable effects on decreasing deception and can contribute to conveying health information to consumers. Text warning labels in their current form have again here shown to be ineffective, similar to previous work (Brennan et al., 2011; Kennedy et al., 2012; Scollo et al., 2015). The results of the current study may also have implications for regulating the marketing and promotion of ”new” tobacco and nicotine-containing products, like reduced-nicotine cigarettes and e-cigarettes. It is essential that policy and public health stakeholders have strong empirical data to inform decisions on how a newly regulated tobacco product (e.g., e-cigarettes, reduced-nicotine cigarettes) can (and cannot) be marketed and labeled to accurately inform consumers about the product. This is a highly timely and relevant topic given the rapidly evolving tobacco product market and the potential to regulate nicotine content in cigarettes.

Highlights.

  • Exposure to a text corrective in print ads decreases false beliefs about the product

  • It improves correct recall of information provided by the corrective

  • Corrective statements in the body of the ad can convey health information to consumers

  • Text-only warnings are insufficiently viewed for conveying information

Acknowledgments

Role of Funding Source

Research reported in this publication was supported by the National Cancer Institute (NCI) of the National Institutes of Health (NIH) and FDA Center for Tobacco Products (CTP) under Award Numbers P50CA179546, and R01CA180929, as well as Robert Wood Johnson Foundation 62622. The content is solely the responsibility of the authors and does not necessarily represent the official views of the NIH or the Food and Drug Administration (FDA). The funding sources had no role in the study design, in the collection, analysis and interpretation of data, in writing the report, or the decision to submit the article for publication.

Footnotes

Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Conflict of interests

No conflict declared.

Contributors

Andrew A. Strasser conceived and supervised the study. Kirsten Lochbuehler completed the analyses. Kirsten Lochbuehler led the writing. Kathy Z. Tang, Valentina Souprountchouk, Dana Campetti, Joseph N. Cappella, and Lynn Kozlowski assisted with the study. All authors have approved the final article.

REFERENCES

  1. Anderson SJ, Pollay RW, Ling PM. Taking ad-Vantage of lax advertising regulation in the USA and Canada: Reassuring and distracting health-concerned smokers. Soc. Sci. Med. 2006;63:1973–1985. doi: 10.1016/j.socscimed.2006.05.020. [DOI] [PubMed] [Google Scholar]
  2. Bansal-Travers M, Hammond D, Smith P, Cummings KM. The impact of cigarette pack design, descriptors, and warning labels on risk perception in the U.S. Am. J. Prev. Med. 2011;40:674–682. doi: 10.1016/j.amepre.2011.01.021. [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. Biener L, Bogen K, Connolly G. Impact of corrective health information on consumers’ perceptions of “reduced exposure” tobacco products. Tob. Control. 2007;16:306–311. doi: 10.1136/tc.2006.019240. [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. Brennan E, Durkin SJ, Cotter T, Harper T, Wakefield MA. Mass media campaigns designed to support new pictorial health warnings on cigarette packets: evidence of a complementary relationship. Tob. Control. 2011;20:412–418. doi: 10.1136/tc.2010.039321. [DOI] [PubMed] [Google Scholar]
  5. de Andrade M, Hastings G, Angus K. Promotion of electronic cigarettes: tobacco marketing reinvented? BMJ. 2013:347. doi: 10.1136/bmj.f7473. [DOI] [PubMed] [Google Scholar]
  6. Donny EC, Denlinger RL, Tidey JW, Koopmeiners JS, Benowitz NL, Vandrey RG, al’Absi M, Carmella SG, Cinciripini PM, Dermody SS. Randomized trial of reduced-nicotine standards for cigarettes. N. Engl. J. Med. 2015;373:1340–1349. doi: 10.1056/NEJMsa1502403. [DOI] [PMC free article] [PubMed] [Google Scholar]
  7. Dunsby J, Bero L. A nicotine delivery device without the nicotine? Tobacco industry development of low nicotine cigarettes. Tob. Control. 2004;13:362–369. doi: 10.1136/tc.2004.007914. [DOI] [PMC free article] [PubMed] [Google Scholar]
  8. Fischer PM, Richards JW, Berman EJ, Krugman DM. Recall and eye tracking study of adolescents viewing tobacco advertisements. JAMA. 1989;261:84–89. [PubMed] [Google Scholar]
  9. Grana RA, Ling PM. “Smoking revolution”: a content analysis of electronic cigarette retail websites. Am. J. Prev. Med. 2014;46:395–403. doi: 10.1016/j.amepre.2013.12.010. [DOI] [PMC free article] [PubMed] [Google Scholar]
  10. Hamilton WL, diStefano Norton G, Ouellette TK, Rhodes WM, Kling R, Connolly GN. Smokers' responses to advertisements for regular and light cigarettes and potential reduced-exposure tobacco products. Nicotine Tob. Res. 2004;6:S353–S362. doi: 10.1080/14622200412331320752. [DOI] [PubMed] [Google Scholar]
  11. Heatherton TF, Kozlowski LT, Frecker RC, Fagerström KO. The Fagerström Test for Nicotine Dependence: a revision of the Fagerström Tolerance Questionnaire. Br. J. Addict. 1991;86:1119–1127. doi: 10.1111/j.1360-0443.1991.tb01879.x. [DOI] [PubMed] [Google Scholar]
  12. Huang J, Kornfield R, Szczypka G, Emery SL. A cross-sectional examination of marketing of electronic cigarettes on Twitter. Tob. Control. 2014;23:iii26–iii30. doi: 10.1136/tobaccocontrol-2014-051551. [DOI] [PMC free article] [PubMed] [Google Scholar]
  13. Kennedy RD, Spafford MM, Behm I, Hammond D, Fong GT, Borland R. Positive impact of Australian ‘blindness’ tobacco warning labels: findings from the ITC four country survey. Clin. Exp. Optom. 2012;95:590–598. doi: 10.1111/j.1444-0938.2012.00789.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. Kozlowski LT, Palmer R, Stine MM, Strasser AA, Yost BA. Persistent effects of a message counter-marketing light cigarettes: results of a randomized controlled trial. Addict. Behav. 2001;26:447–452. doi: 10.1016/s0306-4603(00)00111-8. [DOI] [PubMed] [Google Scholar]
  15. Kozlowski LT, Sweeney CT. Low-yield, light, and ultra-light cigarettes: let's understand the product before we promote. In: Goldberg ME, Fishbein M, Middlestadt S, editors. Social Marketing: Theoretical And Practical Perspectives. Erlbaum and Associates; New York: 1997. [Google Scholar]
  16. Kozlowski LT, Sweeney CT, Palmer RF, Pillitteri JL, White EL, Stine MM, Goldberg ME, Yost BA. Smoker reactions to a “radio message” that light cigarettes are as dangerous as regular cigarettes. Nicotine Tob. Res. 1999;1:67–76. doi: 10.1080/14622299050011171. [DOI] [PubMed] [Google Scholar]
  17. Kozlowski LT, Yost B, Stine MM, Celebucki C. Massachusetts’ advertising against light cigarettes appears to change beliefs and behavior. Am. J. Prev. Med. 2000;18:339–342. doi: 10.1016/s0749-3797(00)00125-2. [DOI] [PubMed] [Google Scholar]
  18. Lohse GL. Consumer eye movement patterns on yellow pages advertising. J. Advert. 1997;26:61–73. [Google Scholar]
  19. Luo C, Zheng X, Zeng DD, Leischow S. Portrayal of electronic cigarettes on YouTube. BMC Public Health. 2014;14:1028. doi: 10.1186/1471-2458-14-1028. [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. Munafò MR, Roberts N, Bauld L, Leonards U. Plain packaging increases visual attention to health warnings on cigarette packs in non-smokers and weekly smokers but not daily smokers. Addiction. 2011;106:1505–1510. doi: 10.1111/j.1360-0443.2011.03430.x. [DOI] [PubMed] [Google Scholar]
  21. National Cancer Institute . Smoking and Tobacco Control Monograph. Vol. 13. National Cancer Institute; Bethesda, MD: 2001. Risks Associated With Smoking Cigarettes With Low Machine-Measured Yields of Tar and Nicotine. [Google Scholar]
  22. O’Connor RJ, Hyland A, Giovino GA, Fong GT, Cummings KM. Smoker awareness of and beliefs about supposedly less-harmful tobacco products. Am. J. Prev. Med. 2005;29:85–90. doi: 10.1016/j.amepre.2005.04.013. [DOI] [PubMed] [Google Scholar]
  23. Paek H-J, Kim S, Hove T, Huh JY. Reduced harm or another gateway to smoking? Source, message, and information characteristics of e-cigarette videos on YouTube. Journal of health communication. 2014;19:545–560. doi: 10.1080/10810730.2013.821560. [DOI] [PubMed] [Google Scholar]
  24. Parascandola M, Augustson E, O’Connell ME, Marcus S. Consumer awareness and attitudes related to new potential reduced-exposure tobacco product brands. Nicotine Tob. Res. 2009;11:886–895. doi: 10.1093/ntr/ntp082. [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Pederson LL, Nelson DE. Literature review and summary of perceptions, attitudes, beliefs, and marketing of potentially reduced exposure products: communication implications. Nicotine Tob. Res. 2007;9:525–534. doi: 10.1080/14622200701239548. [DOI] [PubMed] [Google Scholar]
  26. Morris Co, Bates number 2048718182.
  27. Philip Morris Co, Bates number: 2047387079.
  28. Pollay RW, Dewhirst T. Marketing cigarettes with low machine-measured yields. Smoking And Tobacco Control Monograph. 2001;13:199–233. [Google Scholar]
  29. Pollay RW, Dewhirst T. The dark side of marketing seemingly "Light" cigarettes: successful images and failed fact. Tob. Control. 2002;11:I18–I31. doi: 10.1136/tc.11.suppl_1.i18. [DOI] [PMC free article] [PubMed] [Google Scholar]
  30. Raiff BR, Faix C, Turturici M, Dallery J. Breath carbon monoxide output is affected by speed of emptying the lungs: implications for laboratory and smoking cessation research. Nicotine Tob. Res. 2010;12:834–838. doi: 10.1093/ntr/ntq090. [DOI] [PMC free article] [PubMed] [Google Scholar]
  31. Rayner K. Eye movements in reading and information processing. Psychol. Bull. 1978;85:618–660. [PubMed] [Google Scholar]
  32. Richardson A, Ganz O, Stalgaitis C, Abrams D, Vallone D. Noncombustible tobacco product advertising: how companies are selling the new face of tobacco. Nicotine Tob. Res. 2013;16:606–614. doi: 10.1093/ntr/ntt200. [DOI] [PubMed] [Google Scholar]
  33. Richardson A, Ganz O, Vallone D. Tobacco on the web: surveillance and characterisation of online tobacco and e-cigarette advertising. Tob Control. 2014;24:341–347. doi: 10.1136/tobaccocontrol-2013-051246. [DOI] [PubMed] [Google Scholar]
  34. Scollo M, Lindorff K, Coomber K, Bayly M, Wakefield M. Standardised packaging and new enlarged graphic health warnings for tobacco products in Australia—legislative requirements and implementation of the Tobacco Plain Packaging Act 2011 and the Competition and Consumer (Tobacco) Information Standard. Tob. Control. 2015;2011;24:ii9–ii16. doi: 10.1136/tobaccocontrol-2014-052073. [DOI] [PMC free article] [PubMed] [Google Scholar]
  35. Shadel WG, Lerman C, Cappella J, Strasser AA, Pinto A, Hornik R. Evaluating smokers' reactions to advertising for new lower nicotine quest cigarettes. Psychol. Addict. Behav. 2006;20:80–84. doi: 10.1037/0893-164X.20.1.80. [DOI] [PubMed] [Google Scholar]
  36. Shiffman S, Burton SL, Pillitteri JL, Gitchell JG, Di Marino ME, Sweeney CT, Wardle PA, Koehler GL. Test of “Light” cigarette counter-advertising using a standard test of advertising effectiveness. Tob. Control. 2001;10:i33–i40. doi: 10.1136/tc.10.suppl_1.i33. [DOI] [PMC free article] [PubMed] [Google Scholar]
  37. Shiffman S, Jarvis MJ, Pillitteri JL, Di Marino ME, Gitchell JG, Kemper KE. UK smokers' and ex-smokers' reactions to cigarettes promising reduced risk. Addiction. 2007;102:156–160. doi: 10.1111/j.1360-0443.2006.01650.x. [DOI] [PubMed] [Google Scholar]
  38. Shiffman S, Pillitteri J, Burton S, Di Marino M. Smoker and ex-smoker reactions to cigarettes claiming reduced risk. Tob. Control. 2004;13:78–84. doi: 10.1136/tc.2003.005272. [DOI] [PMC free article] [PubMed] [Google Scholar]
  39. Slade J. Cover essay: the pack as advertisement. Tob. Control. 1997:169–170. doi: 10.1136/tc.6.3.169. [DOI] [PMC free article] [PubMed] [Google Scholar]
  40. Smith P, Bansal-Travers M, O'Connor R, Brown A, Banthin C, Guardino-Colket S, Cummings KM. Correcting over 50 years of tobacco industry misinformation. Am. J. Prev. Med. 2011;40:690–698. doi: 10.1016/j.amepre.2011.01.020. [DOI] [PubMed] [Google Scholar]
  41. Strasser AA, Tang KZ, Romer D, Jepson C, Cappella JN. Graphic warning labels in cigarette advertisements: recall and viewing patterns. Am. J. Prev. Med. 2012;43:41–47. doi: 10.1016/j.amepre.2012.02.026. [DOI] [PMC free article] [PubMed] [Google Scholar]
  42. Strasser AA, Tang KZ, Tuller MD, Cappella JN. PREP advertisement features affect smokers' beliefs regarding potential harm. Tob. Control. 2008;17:I32–I38. doi: 10.1136/tc.2007.022426. [DOI] [PMC free article] [PubMed] [Google Scholar]
  43. Tangari AH, Kees J, Andrews JC, Burton S. Can corrective ad statements based on US v. Philip Morris USA Inc. affect consumer beliefs about smoking? J. Public Policy Marketing. 2010;29:153–169. [Google Scholar]
  44. U.S. District Court DC. U.S. v. Philip Morris USA et al. 2006 I. No. 99-CV-02496GK (Final Opinion) [Google Scholar]
  45. U.S. Food and Drug Administration Center for Tobacco Products. 2015 Accessed July. Issue Snapshot: Modified Risk Tobacco Products (MRTPs). http://www.fda.gov/downloads/TobaccoProducts/Labeling/ProductsIngredientsComponents/Tob.
  46. accoProductReviewEvaluation/UCM455814.pdf.
  47. US Department of Health Human Services Reducing tobacco use: a report of the Surgeon General. US Department of Health and Human Services. 2000 [Google Scholar]
  48. Wakefield M, Morley C, Horan JK, Cummings KM. The cigarette pack as image: new evidence from tobacco industry documents. Tob. Control. 2002;11:i73–i80. doi: 10.1136/tc.11.suppl_1.i73. [DOI] [PMC free article] [PubMed] [Google Scholar]

RESOURCES