Building Resistance to Attitude Change Toward a Health

Building Resistance to Attitude Change Toward a Health Product David Gadiuta Roger Marshall Drew Franklin April 2016

Roger Marshall ý Professor of Marketing 2

The School of Business at Auckland University of Technology 3

The School of Business at Auckland University of Technology

Roger Marshall ý Professor of Marketing ý A Kiwi returned home ý A business background ý Undergraduate economics major, psychology minor ý Doctorate from UWA in cognitive psychology and consumer behaviour ý Teaching, leadership, and 5 research, research…

① Brief background i. Attitude formation and change ii. Resistance to attitude change iii. Research objectives Outline ② The research conducted i. Overall design ii. Respondents iii. Research instrument & process iv. Results ③ Rounding it all up i. Value to theory ii. Value to practice iii. Where to now?

Message Interested in topic? Interested in message? NO Cue-driven peripheral route YES Consider the arguments Cacciopo & Petty’s (1986) attitude change model (stylised) Argument strength Persuasion

Reasons for wanting to confer resistance…

Reasons for wanting to confer resistance… Techniques to confer resistance (mostly via the military) Traditional psychological techniques: Personality, motivational states Training in logical analysis Public commitment Anchoring effect (from Balance Theory)

Reasons for wanting to confer resistance… Techniques to confer resistance (mostly via the military) Traditional psychological techniques: Personality, motivational states Training in logical analysis Public commitment Anchoring effect (from Balance Theory) The immunization analogy (Mc. Guire, 1969)

16 14 The Immunization analogy, (Mc. Guire & Papageogius, 1961) 12 10 8 6 4 2 0 Original No Supportive Source Refutational Peer attitude defense disagreement derogation

The research purpose P&C suggest that a strong argument is most effective in forming (or changing) attitude. Others have cast doubt on this idea, or ammended it, to suggest that the effect of strong and weak arguments vary over time (Bither et al 1971; Eisand 2006) So the research question addressed here is “Is a weak persuasive message more effective in conferring resistance to attitude change over time than a strong message? ” We chose a health-related frame as this tends to be high involvement…

The experiment: All subjects (136) exposed to a scenario: Smokers Toothpaste – old vs. new Original brand is attacked by the new competitor, claiming faster results with less frequent use Manipulation introduced: Inoculation treatment GP 1 (38. 2%) CONTROL GP 2 (38. 2%) STRONG GP 3 (23. 5%) WEAK MEASURE ATTITUDES Allow 14 days to pass Re-expose to base scenario (sans treatment) MEASURE ATTITUDES

Sample Characteristics ü 404 subjects initially, only 136 used due to drop out between time periods ü North American region – 18% smokers (cancer. org), ease of access ü Self identified smokers (min 20 cigarettes per week) ü 18 to 55 age range ü 50/50 Male to female split (68 each) Sample Splits q Age q Smoking regularity q Gender q Education

Process o o Undergrad students for preliminary pre testing (30) Used 7 point Likert scales throughout Survey information sheet + experiment explained Self completion questionnaire (Cint panel service)

Process o o Undergrad students for preliminary pre testing (30) Used 7 point Likert scales throughout Survey information sheet + experiment explained Self completion questionnaire (Cint panel service) Scenarios

Process Undergrad students for preliminary pre testing (30) Used 7 point Likert scales throughout Survey information sheet + experiment explained Self completion questionnaire (Cint panel service) Face validity Post hoc identification of ambiguous statements Clean the data Factor analysis – scales all load cleanly Scale reliability, all scales well over 0. 70 for both period 1 and period 2 o Hypothesis testing: T-tests, one-way ANOVA, Tukey’s HSD o o o o o

Results ² Main effect is positive ² Strong argument works best in the short term ² Weak counter argument =/> no argument Loyalty Over Time 5. 0 4. 5 4. 0 3. 5 3. 0 2. 5 2. 0 1. 5 1. 0 0. 5 0. 0 Time A Time B Control 3. 6 Strong 4. 5 3. 4 Weak 3. 9 4. 0

Results ² Main effect is positive ² Strong argument works best in the short term ² Weak counter argument =/> no argument Loyalty Over Time 5. 0 4. 5 4. 0 3. 5 3. 0 t = 4. 2 , p =. 001 2. 5 t = 2. 2, p =. 03 2. 0 1. 5 t = 2. 2, p =. 035 1. 0 0. 5 0. 0 Time A Time B Control 3. 6 Strong 4. 5 3. 4 Weak 3. 9 4. 0

Possible moderators u Gender differences are not significant u Age differences also not supported, but probably due to the small number of young smokers (as there is a quite large absolute difference) u I broke up “Loyalty” to cognitive, affective and behavioral components, and “went fishing” to see which is of primary importance. All have an effect: - cognitive (t = 2. 5, p=. 011 - emotion (t = 2. 3, p =. 01 - behavior (t = 2. 9, p =. 003). Interesting to speculate (but it is speculation only!) that the strong behavioral component might imply a sub-conscious response and not just a conscious cognitive reasoning process?

Possible moderators, behavior & smoker frequency Low Frequency Smoker Loyalty Over Time 5. 5 5. 0 4. 5 Means 4. 0 3. 5 3. 0 2. 5 2. 0 Control (1) LF SMKR Strong (2) LF SMKR Weak (3) LF SMKR Time A 3. 5 4. 6 3. 9 Time B 3. 7 3. 0 5. 3

Possible moderators, behavior & smoker frequency High Frequency Smoker Loyalty Over Time 4. 5 Note: Observed differences are not statistically significant 4. 0 Means 3. 5 3. 0 2. 5 2. 0 Control (1) HF SMKR Strong (2) HF SMKR Weak (3) HF SMKR Time A 4. 0 3. 9 Time B 3. 5 3. 3 3. 7 Axis Title

Further research q Clearly three stages better than two to get a trend q Re-arrange data or collect more to test the age hypothesis that did seem to want to be significant on the very small cell size q Mechanisms – what causes the effect? Mediation and moderation using various factors might help q The role of WOM in innoculation – the digital connection

Implications for theory v Confirmatory; supports the ELM, Mc. Guire and Eisand – i. e. , there does appear to be an internal debate going on v & there is a time dimension that affects attitude change v This is the first research that uses a longitudinal design in a marketing context, and the first to extend the time between exposures for more a few days Implications for marketing It’s worthwhile attempting to stop attitudes changing! Not only is it cheaper to retain clients than find new ones, but it is also cost effective in that a single message can act as both an inoculation and as an attitude change device

Implications for theory v Confirmatory; supports the ELM, Mc. Guire and Eisand – i. e. , there does appear to be an internal debate going on v & there is a time dimension that affects attitude change Implication for v This is the first researchme that uses a longitudinal design in a marketing context, and the first to extend the time between Hopefully, exposures for more. I a few days Implications for marketingget a It’s worthwhile attempting to stop attitudes changing! publication! Not only is it cheaper to retain clients than find new ones, but it is also cost effective in that a single message can act as both an inoculation and as an attitude change device

Survey Questions & Scenarios: Base scenario As a smoker consciously looking after your dental hygiene, consider that for several years, you have been using a toothpaste brand especially for smokers, named “Crown”. This specially formulated toothpaste aids you in countering the negative discoloring effects on teeth caused by smoking. Throughout your use of the Crown brand, you have not experienced any side effects nor any problems. The whitening treatment it promises has been generally effective. With frequent use of the Crown toothpaste, you are able to keep the attractive white coloring of your teeth. While doing your shopping and seeking out your regular smoker’s toothpaste, you notice a new competing brand ‘Royal, ’ which is selling for the same price as your regular brand. You recall having seen advertising from Royal, which claimed to act much faster and stronger than any existing brand. Thanks to its speedy results, the new Royal brand claims that you would even be able to reduce the treatment frequency and amount of time spent brushing.

Strong counter-argument version: As you are now considering the decision of which brand to purchase, you remember seeing advertising from your regular smoker’s toothpaste, Crown. Their advertisement claims that new competitors (such as Royal), only achieve their quick results through the use of a chemical that is proven to cause tooth decay, thus achieving only temporary cosmetic effects. According to Crown, the Royal smoker’s toothpaste product fails to aid in the long term improvement of your oral hygiene, and puts your teeth at risk. Weak counter-argument: As you are now considering the decision of which brand to purchase, you remember seeing advertising from your regular smoker’s toothpaste, Crown. The advertisement highlights the fact that they are very experienced at making smokers’ toothpaste, unlike newer market entries. Crown implores you to stick with the brand you know and trust.

Cognitive questions: Given the effective results I have experienced using Crown, I have no reason to consider a new toothpaste. I have given serious consideration to the claims of Royal, the new smoker’s toothpaste. I think that the new Royal brand toothpaste offers better value than the Crown toothpaste I currently use. Emotional questions: I would be excited to switch to Royal, the new fast acting smokers toothpaste. I feel very attached to Crown, my current and effective smokers tooth paste. I’m uncomfortable with the idea of switching from Crown to the new brand, Royal. Purchase likelihood questions: I will stick with my regular smoker’s toothpaste, and continue to buy Crown. I am keen to purchase the new smoker’s brand, Royal. The next time I buy smokers toothpaste I will ignore the new brand, Royal.

Back to experiment design