The usage of subliminal messages in advertising: an empirical investigation about their efficacy
1 like2,083 views
This study investigated the efficacy of subliminal messages in print advertising through a mixed methods approach. A qualitative phase involved interviews to understand perceptions of advertising, followed by a quantitative survey with 267 respondents randomly assigned to control and test groups evaluating ads with or without embedded sexual images. For a bacon cheeseburger ad, the test group rated the ad and brand more positively on several measures. However, for a gin ad, only one measure ("insignificant to memorable") differed between groups. The author concluded subliminal messages may influence some product categories more than others and that message type and product category need further investigation to understand how to best apply subliminal techniques. Limitations included a small sample size and limited message/product types
The usage of subliminal messages in advertising: an empirical investigation about their efficacy
- 1. The usage of subliminal messages in advertising: an empirical investigation about their efficacy MSc in Marketing Management Academic Year 2015-2016 Antonino Ciampa
- 2. Agenda 2 Research objectives2 Methodology Results4 Results’ discussion and implications5 Limitations and further research6 Executivesummary1
- 3. Research objectives 4 Hypotheses The inclusion of an embed material in print advertising will have a significant impact on one or more measures of: 1 2 3 4 • First impression (open question)
- 4. Methodology 5 The empirical research was structured into the following steps: DESK RESEARCH Insights from other studies about the structure and method of the research QUALITATIVE PART 9 in-depth interviews plus an extra interview with an advertising agent QUANTITATIVE PART Data collected through a survey, with 267 responses. ANALYSIS Results analyzed with SPSS, to verify the validity of the hypotheses
- 5. Methodology – Desk Research 6 Conflicting results: • Kilbourne, Painton, and Ridley (1985) demonstrated the positive effects of subliminal messages on a liquor print ad. • Rosen and Singh (1992) did not find any statistically significant effects for embeds at any level of advertising effectiveness they measured, for liquor and cologne ads. Wilson Bryan Key has emphasized the supposed use of embeds in advertising. In Subliminal Seduction, he claimed that advertisers embedded images of sex, death, and bestiality in their ads to manipulate people into buying their products.
- 6. Methodology – Qualitative Research 7 9 In-depth interviews - 6 women, 3 men - Respondents between 21 and 45 years old Food and high-alcohol-content beverage product categories chosen Colors and subjects in a print ad are more likely to catch attention 10 seconds: average exposition time to print ads 1 extra in-depth interview with an advertising agent
- 7. Methodology – Quantitative Research 8 67% 20% 4% 3% 4% 2% 18-24 years old 25-34 years old 35-44 years od 45-54 years old 55-64 years old 65 years old or more 51%49% Male Female 267 respondents 2 control ads and 2 test ads -all groups homogenous in terms of age, gender and occupation 62% believe in subliminal messages’ existence 85% believe subliminal messages influence choices Do you believe in the existence of subliminal messages? (n=244) 83% 17% Italian Other of which
- 8. Methodology – Quantitative Research 9 Test group (126 respondents) Original advertising Control Group (136 respondents) Modified advertising Respondents evaluated a gin ad and a bacon cheeseburger ad Test ads contained similar sexual embed Several tests conducted to verify that the messages were “subliminal” Respondents were randomly assigned to groups No subject evaluated both versions of an ad
- 9. Methodology – Quantitative Research 10 Control Group (115 respondents) Original advertising Test Group (135 respondents) Modified advertising
- 10. Results – Gin ad 11 Evaluation of the ad •Insignificant/Memorable (p=.065) •No other significant differences Brand favorability •No significant findings Intention to buy •No significant differences Willingness to pay •No significant differences Open question elaboration •Confirm what emerged from the statistical analysis •The test group had better impressions Gin ad Boring/ Interesting Insignificant/ Memorable Unpleasant/ Pleasant Overall liking NPS Control Group Mean 5,14 4,60 5,97 5,46 3,79 N 136 136 136 136 136 Std. Deviation 2,294 2,019 2,157 2,011 2,226 Test Group Mean 5,56 5,06 6,33 5,64 4,13 N 126 126 126 126 126 Std. Deviation 2,006 2,003 2,032 1,899 2,117 The table below shows ratings of items for the evaluation of the ad. Ratings for the test group are greater than control group’s, for the following items:
- 11. Results – Bacon Cheeseburger ad 12 Evaluation of the ad • Boring/Interesting (p=.029) • Insignificant-Memorable (p=.065) Brand favorability • Unpleasant/Pleasant (p=.04) • Significant difference at 0.1 level for all the other items Intention to buy • Delta intention to buy (p=.065) Willingness to pay • No significant differences Open question elaboration • Confirm what emerged from the statistical analysis • The test group had better impressions on brand and ad’s evaluations ANOVA Table Brand attitudes Sum of Squares df Mean Square F Sig. Worthless/ Valuable Between Groups 12,824 1 12,824 3,879 ,050 Within Groups 803,397 243 3,306 Total 816,220 244 Unattractive/ Attractive Between Groups 10,386 1 10,386 3,647 ,057 Within Groups 691,924 243 2,847 Total 702,310 244 Negative/ Positive Between Groups 9,133 1 9,133 3,016 ,084 Within Groups 735,863 243 3,028 Total 744,996 244 Unpleasant/ Pleasant Between Groups 14,241 1 14,241 4,267 ,040 Within Groups 811,040 243 3,338 Total 825,282 244
- 12. The difference between control and test group is: Results – Wrap up 13 Gin ad Bacon Cheeseburger ad Evaluation of the ad Boring/Interesting Uninformative/Informative Insignificant/Memorable Unpleasant/Pleasant Overall liking Net Promoter Score Brand favorability Worthless/Valuable Unattractive/Attractive Negative/Positive Unpleasant/Pleasant Intention to buy Intention to buy Delta intention to buy Willingness to pay Willingness to pay Delta willingness to pay Significant at 0.05 level Significant at 0.1 level Not significant
- 13. Results’ discussion and implications 14 Suggested process: Evaluation of several versions of the same advertising with different types of subliminal message through a quantitative research to check which type of embeds should be used. Dissimilar impact of subliminal messages on product categories: • Positive influence on certain variables in cheeseburger ad • Limited influence in gin ad Studies taken as a whole have produced a mixture of conflicting results Positive results are not generalizable to all products Generic managerial implications might not be reliable
- 14. Limitations and further research 15 Effectiveness of subliminal messages might be more dependent on the type and form of embed used and the product involved Additional study should investigate the relations among them to understand which kind of subliminal message better fit a specific product category. - Small sample - The sample is not an adequate representation of the population (majority belonging to the age range 18-24) Limited product categories covered Limited usage of subliminal messages’ types Lack of a realistic advertising presentation environment
Editor's Notes
- #4: The inclusion of an embed material in print advertising will have a significant impact on one or more measures of the advertisement’s effectiveness. The inclusion of an embed material in print advertising will have a significant impact on one or more measures of brand favorability. The inclusion of an embed material in print advertising will have a significant impact on respondents’ intention to buy. The inclusion of an embed material in print advertising will have a significant impact on respondents’ willingness to pay.
- #6: Several authors have attempted to clarify the phenomenon in the past. and still today there is no scientific evidence about subliminal messages’ efficacy. Critical for defining: Objectives and hypotheses of the study Structure of the study, by focusing on the differences between a control group and a test group Scales and questions for the questionnaire
- #7: During this phase, product categories were defined according to these criteria: Familiarity of the product category Often-purchased product category The product categories chosen for the investigation were food and high-alcohol-content beverage
- #8: Respondents were asked to evaluate a gin ad and a bacon cheeseburger ad*. From these two experimental ads, two control ads and two test ads were developed and respondents were randomly assigned to groups and no subject received both versions of an ad Attitudes towards the print ad: Boring/Interesting , Uninformative/Informative, Insignificant/Memorable, Unpleasant/Pleasant. Additionally, were also assessed the overall evaluation of the ad and the Net Promoter Score of the product publicized. Brand attitudes: Worthless/Valuable, Unattractive/Attractive, Negative/Positive, Unpleasant/Pleasant. Open questions were also used to analyse further differences between control and test group. Finally, other questions aimed to investigate respondent’s intention to buy and willingness to pay.
- #11: In the gin ad the subliminal message did not seem to be effective. Results show that: Evaluation of the ad. No significant differences between control group and test group. The only exception is the measure Insignificant-Memorable (p=.065). As expected, the test group rankings for all evaluations (apart for Uninformative-Informative) are greater than those of the control group. Brand attitudes. No significant findings arose from the differences of evaluations. Intention to buy. No significant difference emerged relatively to intention to buy and delta intention to buy (difference between intention to buy of the product publicized and intention to buy for the generic product) Willingness to pay. No significant difference for willingness to pay and delta willingness to pay Results from the open questions elaboration confirm what emerged from the statistical analysis, where the test group had on average higher evaluations.
- #12: In the gin ad the subliminal message did not seem to be effective. Results show that: Evaluation of the ad. No significant differences between control group and test group. The only exception is the measure Insignificant-Memorable (p=.065). As expected, the test group rankings for all evaluations (apart for Uninformative-Informative) are greater than those of the control group. Brand attitudes. No significant findings arose from the differences of evaluations. Intention to buy. No significant difference emerged relatively to intention to buy and delta intention to buy (difference between intention to buy of the product publicized and intention to buy for the generic product) Willingness to pay. No significant difference for willingness to pay and delta willingness to pay Results from the open questions elaboration confirm what emerged from the statistical analysis, where the test group had on average higher evaluations.