Healthcare Recommendations:
The influence of ego-involvement on ad-liking, brand attitude and referral intentions when resistance and personal-relevance mediate.
Laura M. Pyter - 11375922 Master’s Thesis
Graduate School of Communication
Master’s programme in Communication Science University of Amsterdam,
Supervised by Eva van Reijmersdal
Abstract
This study examines the hypothesis that framing ego-involvement in advertising influences (a) ad-liking, (b) brand attitude and (c) referral intentions, when mediated by resistance and relevance. It also suggests that differences between the high and the low ego-involvement groups exist when measuring resistance and relevance; respectively. One hundred healthcare practitioners were exposed to a fictitious advertisement under high and low ego-involvement conditions and subsequently their resistance, relevance, ad-liking, brand attitude, and referral intentions were assessed. The results do not support the hypothesis that an indirect effect took place and both groups were found to be more similar in nature than different when measuring resistance and relevance levels. However, the positive correlation between relevance and (a) ad-liking, (b) brand attitude and (c) referral intentions was significant. Findings are discussed in terms of the influence of the ego-involvement construct that results in conflicting measures of resistance, relevance and in turn (a) ad-liking, (b) brand attitude and (c) referral intentions. Limitations and suggestions for future research are addressed.
Introduction
The ongoing exploration of what does and does not work in advertising development, spans across many fields interested in different aspects of the process. From a communication standpoint, there has been an interest in defining the involvement construct as it has appeared in many studies which investigate the effects of framing messages on consumer behavior within advertising. This current study builds upon previous research which has explored the antecedents and consequences of involvement and the varying types that exist. The four streams of research within involvement include: attention, situational, audience and enduring (Muehling & Laczniak, 1988). However, researchers go on to use the terms interchangeably and tend to overlap existing measures by defining terms in various ways. The effects of involvement have been examined with other factors such as mental processing, resistance, attitude formation, etc. Simply put, involvement can focus on a temporary or long-lasting relation to a topic, or look at external factors influencing what causes a person to be involved. It is has been also been established that consumers use specific resistance strategies towards persuasive messages in various contexts (Jacks and Cameron, 2010). This study focuses on the relation between involvement with regards to research that addresses resistance to persuasive messages. More specifically, ego-involvement (a form of enduring ego-involvement) is explored. Ego-involved messages integrate important ego factors within a task, “e.g., social prestige, self-esteem, fear of academic
standing”, causing the subject to feel that they are directly involved with it (Klein & Schoenfeld, 1941, p. 249). Ego-involved messages, then, will be defined as messages which highlights importance to one’s self (high ego-involvement) or the lack there of (low ego-involvement). Researchers and marketers focused on the healthcare sector have looked at consumer responses to advertising efforts. However, as consumers are not regarded as experts when advising in
medical product recommendations, this study will strictly observe the healthcare practitioner’s response to ego-involved persuasive messages. Healthcare practitioners include all medical professionals who have acquired the expertise and authority to influence and recommend health-related products to consumers (i.e. doctors, nurses and other medical staff). As the motivational consequence in a high ego-involvement framed message is the attachment of relational
importance one has to the situation, activation of a relatively high mental effort is seen (Gendolla, Brinkmann & Scheder, 2008). Therefore, with the exploration of ego-involvement effects on ad-liking, brand attitude and referral intentions it will be possible to measure healthcare practitioners’ resistance and susceptibility to ego-framed persuasive messages.
The research question in consideration is as follows: In the healthcare context, what is
the influence of ego-involvement based persuasive messages on (a) ad-liking, (b) brand attitude and (c) referral-intentions and is it mediated by resistance and relevance? In turn, by identifying
the influence of the ego-involvement construct in advertising, suggestions for subsequent development of persuasive messages will be discussed. Theoretical implications suggest that in the context of healthcare communication, ego-involvement may play a role in attitude formation (i.e. ad-liking, brand attitudes, referral intentions). This builds upon previous research which aims to uncover the motivations and uses of resistance in various settings. Practical implications on the other hand provide health-marketers with a better understanding of developing successful persuasive messages in a healthcare context.
Theoretical Background Persuasive messages in healthcare
Healthcare practitioners have generally been the gateway influencers for those looking for advice and guidance on health-related services (Zwier, 2017). However, with interest of
health-related companies in persuading and providing attractive offers to general practitioners, it is difficult to be certain that product referrals are recommended to consumers with the right intentions (Snyder, 2011). Marketing has become an integral part of health organizations worldwide, but it is still being determined how healthcare practitioners respond to advertising (Yavas & Riecken, 2001).
In the healthcare advertising context, healthcare practitioners are seen as the experts while general consumers can be thought of as a “naive public” (Huh & Langteau, 2007). Experts are defined as people who have developed a substantial amount of knowledge on a particular topic, through repeated practice; hence the presence of healthcare practitioners acting as the experts in this particular context. A majority of healthcare practitioners have spoken out on favoring stricter regulation of direct advertising to consumers, given that consumers may be seen as vulnerable, but limited evidence is available on a practitioner’s vulnerability to health-related advertising (Huh & Langteau, 2007). The American Medical Association policy recommends that healthcare practitioners should be monitoring health-related advertising and report to the FDA of any advertisement that fails to provide enough information to patients (Riddick, F., 2003). In this sense, healthcare practitioners have an enduring (ego) involvement with health-related advertising.
The involvement construct
There has been inconsistencies on how to define and measure the construct of involvement in advertising which is likely due to the different applications of the term
“involvement” (Zaichkowsky, 1994). Various types of involvement have been measured: ego, product, brand, enduring, response, situation, purchase, etc. The type of involvement discussed further relates to enduring involvement which has been defined as an ongoing concern, as
compared to situational involvement which reduces one’s involvement to a temporary or special condition (Klein & Schoenfeld, 1941). Greenwald (1982) further define’s ego-involvement as pertaining to one’s self-concern and personal importance.
Ego-involvement & resistance
Persuasive messages are formed with the intent of achieving a change in one’s attitude (Gas, 2015). In order to counter these messages, people engage in different resistance strategies which may result in successful resistance ranging from a no-attitude change to a boomerang effect (an attitude change in the opposite direction than as intended) (Jenkins & Dragojevic, 2011). Jacks and Cameron (2010, p.148) identified seven different resistance strategies according to consumer’s self-reports: counter-arguing, attitude bolstering, social validation, selective exposure, negative affect, source derogation and assertion of confidence. However, in the case of disclosures to sponsored content, participants who were explicitly told they were provided with was sponsored content had used both counter-arguing and negative affect as resistance strategies to overcome any persuasive influence (van Reijmersdal et al., 2016). In effect, attitudes towards the brand had become more negative and purchase intentions decreased. Counter-arguing does not only include direct rebuttals of message arguments but rather extends into contesting the message source or persuasive strategy itself (Fransen, Smit, & Verlegh, 2015). For example, when one is aware of a message’s persuasive intent, their
knowledge of persuasion can (in a sense) protect them from any unauthorized influence attempt on their existing position by using a counter-response to the message itself (Pfau et al., 1997). In the case of presenting a health-related persuasive message as an advertisement, counter-arguing is measured as the prevalent response. The effect of resistance on brand attitudes and ad-liking
has been found to be dependent on many factors, ego-involvement being one (Halverson & Pallak, 1978; Pfau et al., 1997).
The conceptual model shown (Figure 1) proposes a relationship between
ego-involvement and resistance and includes another pathway through relevance, in turn leading to: ad-liking, brand attitudes and referral intentions (further discussed).
Ego-involvement integrates important ego factors within a task, “e.g., social prestige, self-esteem, fear of academic standing”, causing the subject to feel that they are directly involved with subsequent performance outcomes (Klein & Schoenfeld, 1941, p. 249). In this light, studies have used tests to manipulate ego-involvement by tying self-esteem to a performance outcome. Advertisements or other persuasive messages which are framed with higher
situational-involvement led to a higher exertion of mental effort when processing messages (Gendolla, Brinkmann & Scheder, 2008). Another experiment on ego-involvement was conducted using mental effort mobilization. In order to test for expectations against ego-involvement, neutral stimuli were used. This study provided evidence of the effects of ego-involvement on general
attitudes, indicating that high ego-involved messages resulted in higher mental effort as
evaluation time grew significantly longer (Gendolla & Richter, 2006). This, in turn, would have a lasting effect on final attitude formation. In observations of attitude-consistent behavior, high ego-involvement subjects were found to be more resistant to attacks on their position and were more likely to take action as compared to low ego-involvement subjects (Halverson & Pallak, 1978). However, it was also found that higher ego-involvement had ultimately facilitated susceptibility to the influence of persuasive attempts (Pfau et al., 1997). More evidence must be addressed in the health-related context on healthcare practitioners and their susceptibility to ego-involved messages. This leads to the first proposed hypothesis:
H1: Healthcare practitioners exposed to a high ego-involvement message show lower levels of resistance as compared to those exposed to a low ego-involvement message.
In the present study, ego-involved messages are defined as messages which vary in the degrees of relation to one’s self and status feature (low to high ego-involvement). Based on literature showing the influence of ego-involvement on attitude formation, it is predicted that high ego-involved messages will result in lower activation of resistance processes. With higher levels of resistance, there are less chances of ad-liking, brand attitude or positive intentions to refer a product which leads to the next hypothesis:
H2: Resistance to persuasion is negatively correlated with (a) ad-liking, (b) brand attitude and (c) referral intentions.
As seen with literature observing the effects of disclosures, attitudes towards the brand had become more negative and purchase intentions had decreased therefore showing that resistance had been activated (van Reijmersdal et al., 2016). The higher resistance one
experiences, the more likely it will be that ad-liking and brand attitudes are lowered and referral intentions to the product would not be recommended. In turn, it is hypothesized that the
influence of ego-involvement on (a) ad-liking, (b) brand attitudes and (c) referral intentions intent to refer a product is mediated by resistance:
H3: The influence of ego-involvement on (a) ad-liking, (b) brand attitude and (c) referral intention is mediated by resistance to persuasion.
In the second hypothesis, the effects of resistance on final brand attitude, ad-liking and referral intention was addressed, but the effect of whether a high or low ego-involved message was not in focus. In this final hypothesis, high ego-involved messages should be seen as the most likely to induce a higher resistance to the persuasive attempt and in turn lower ad-liking, brand attitudes and intent to refer the product.
Ego-involvement and Relevance
Ego-involvement’s relationship to relevance has been explored in terms of personal attachment and goal one has with a topic (Klein, 1941). For example, if an individual’s goal while viewing an advertisement may be to buy the product, the advertisement would be deemed as highly relevant. Otherwise, those who would never buy the advertised product would consider it irrelevant. As ego-involvement relates to a personal connection between one’s self and the object in focus, and ego-involvement has been used in terms of measuring the level of relevance,
the level of ego-involvement and relevance should go hand in hand. However, the difference between relevance and ego-involvement is that relevance can extend outside one’s self while ego-involvement is a direct attachment of the self to an object (Madden, Allen, & Twible, 1988). The first hypothesis is as follows:
H4: Personal relevance reported is higher with exposure to high ego-involvement messages as compared with exposure to low ego-involvement messages.
Personal relevance has been established as having an influence on subsequent attitudes towards a product (Madden, Allen, & Twible, 1988). In fact, brand attitudes are higher when personal relevance is also high (Muehling & Laczniak, 1988). These studies have established the next hypothesis:
H5: Personal relevance is positively correlated to (a) ad-liking, (b) brand attitude and (c) referral intentions.
It was found that when persuasive messages were irrelevant, less favorable consumer outcomes were likely to occur (Yagci, Biswas & Dutta, 2009). This led to the final hypothesis that relevance will mediate the influence of ego-involvement:
H6: The influence of ego-involvement on (a) ad-liking, (b) brand attitude and (c) referral intention is mediated by personal relevance.
Method Subjects
A purposive convenience sample of healthcare practitioners (defined as: doctors, nurses, clinical scientists and PhD experts in the healthcare field) was gathered to complete a cross-sectional online survey. Of the 399 participants that the survey was distributed to, only those which identified themselves as experts in the field were allowed complete the survey in its entirety. After removing incomplete entries, 100 participants were accounted for; 77% of the final sample being female, 23% reported as male. The mean age fell within the 25-34 year old group, with an average of 5-10 years of experience in the healthcare profession. The selection of healthcare practitioners were gathered from Global Survey (i.e. a firm specializing in panel recruitment for healthcare practitioners all over the world). Listed professions included but were not limited to: registered nurses, doctors, physicians assistants, psychologists, therapists and dietitians within the United States.
Design and procedure
An experiment with a one-factor (involvement message versus message with no ego-involvement), between-subjects design was used. The measurement of all variables were
recorded at one point in time, after the manipulation occurred. Participants were randomly assigned to one of the two conditions, allowing all individuals to have equal chances of being in either condition, while avoiding any selection bias. Anonymity was stressed before and after the completion of the survey. Before the survey was launched, it was pretested for logic, flow and timing to ensure participants correctly understood the questions and were able to complete it within a reasonable time. The pretest consisted of 20 healthcare practitioners (not included in the final survey sample). The effectiveness of the ego-involvement manipulation was assessed by
asking participants to indicate how much they felt their response affected the promotion of the product, on a 7point scale (ranging from 1= strongly disagree to 7= strongly agree). Changes were made to the ego-involving message to strengthen the high and low ego-involvement manipulation. Brand manipulation check used two items on a 1-5 scale (α = .92, M = 3.12, SD =.08). The survey preparation and data collection took place from 10 December 2017 to 10 January 2018.
Stimulus materials
Participants received a link to the online experimental survey that initially provided a fact sheet and consent form. It was designed to allow the participant to continue on the condition that they identified themselves as a healthcare practitioners. The survey then featured an
advertisement that included an ego-involved persuasive message (at the top of the page) for the high ego-involvement group, or displayed the advertisement on its own (with no ego-involved message) for the low ego-involvement group. Ego-Involvement includes situations in which important ego factors like “social prestige, self-esteem, fear of academic standing” etc. are tied into the message (Klein & Schoenfeld 1941, p. 249). Addressing the participant as an expert in their field was used as a form of inducing ego-involvement (see Appendix A for manipulation). The low ego-involvement persuasive message did not include any reference to one’s ego factors. The print advertisement was created with a pain-reliever medication as the product in focus (see Appendix D for advertisement). To reduce influence from existing attitudes, a fictitious brand name was used. The advertisement displayed two images of the pain reliever medication; one image of the medication bottle and the other image of the boxed-packaging of the medication. Besides receiving an alternating ego-involvement message, both conditions received the same advertisement as described. The last portion of the survey asked participants to answer questions
in response to the advertisement (i.e. their experienced resistance, how relevant it was, their attitudes towards ad and brand, their intentions for referring the product, etc.). The survey ended with a request for demographics and then participants were debriefed. The exact manipulation and measures of the survey can be found in Appendix A-C.
Measures
Resistance. Resistance in the form of counter-arguing is measured using a four-item,
seven-point scale (1= “strongly disagree”; 7= “strongly agree”). The scale adopted from Fransen, Ter Hoeven, and Verlegh (2013) and Zuwerink & Cameron (2003), asked the participant to indicate include whether one “contested”, “refuted”, “doubted”, or “countered” the information in the message. Mean scores were calculated to create a single measure of resistance (α = .799,
M = 3.06, SD =1.16).
Relevance. The personal relevance of the advertisement is measured using
Zaichkowsky’s (1994) 10-item, Personal Involvement Inventory. This is a semantic differential scale, ranging from 1 (strongly disagree) to 7 (strongly agree), includes items which are reverse scored (see Appendix C). The scale asks how “important”, “interesting”, “relevant”, “exciting”, “meaningful”, “appealing”, “fascinating”, “valuable”, “involving” or how “needed” the product is; as perceived by the participant. Mean scores were calculated to create a single measure of relevance (α =.803, M = 5.05, SD =.89).
Ad Liking. A five-item Likert scale is used to measure ad liking. The five items, adopted
from MacKenzie, Lutz, and Belch (1986) and Madden, Allen, and Twible (1988), are:
“interesting,” “good,” “likable,” “favorable,” and “pleasant” ranging on a 5-point scale from 1= “strongly disagree” to 5= “strongly agree”. Mean scores were calculated to create a single measure of ad liking (α = .902, M = 3.58, SD =.78).
Brand Attitude. Ajzen and Fishbein (1980)’s three item, bipolar scale for brand attitude
includes: “bad/good”, “unfavorable/favorable” and “negative/positive” ranging on a 5-point scale from 1= “strongly disagree” to 5= “strongly agree”. Mean scores were calculated to create a single measure of brand attitude (α = .956, M = 5.63, SD =1.37).
Referral-Intentions. Referral-intentions are measured using two Likert-scale type
questions. The questions, “How likely are you to recommend [product] to someone else?” and “How likely are you to refer [product] to a patient?”, are measured using a 7-point scale (1= “extremely unlikely”, 7= “extremely likely”). Mean scores were calculated to create a single measure of referral-intention (α = .862, M = 5.01, SD = 1.32).
Results
Analysis of variance showed that participants in the two conditions did not differ with respect to sex, x2(1) = .054, p = .59, or age, F(1,98) = .547, p = .46, indicating successful randomization. A quarter of the sample (25 participants) claimed to have used or were aware of the branded medication that was displayed within the advertisement. The claims were almost exactly even between the two conditions. A manipulation check confirmed that no major discrepancies were found amongst the two groups (F(98)= 2.65, p= .107) and the manipulation (see Appendix for all measures) was successful; (t(98) = -2.56, p = .012). There was no
significant difference in scores for the low ego-involvement group (M= 2.73, SD= .95) and the high ego-involvement group (M= 3.22, SD= .92); t(98)= -2.56 p= .012, two-tailed. One thing that must be noted is that the sample differs by one person for results with regards to the measure of referral intention (N=99).
According to hypothesis 1, the group exposed to the high ego-involvement message should experience an overall higher level of resistance as compared to those exposed to to a low
ego-involvement message. An independent samples t-test was performed to compare the
resistance levels experienced between the high involvement group (N= 51) and the low involvement group (N= 49). As can be seen in Table 1, the high involvement and low ego-involvement distributions were sufficiently normal for the purpose of conducting a t-test (i.e.,
skew < |2.0| and kurtosis < |9.0|; Schmider, Ziegler, Danay, Beyer, & Bühner, 2010). Additionally, the assumption of homogeneity of variances was tested and satisfied via Levene’s F test, (F(98)= 2.65, p= .107). There was no significant difference in scores for the low ego-involvement group (M= 3.20, SD= 1.31) and the high ego-involvement group (M= 2.92,
SD= .99); t(98)= 1.23 p= .221,
two-tailed. These results suggest that overall, resistance levels were generally low for both groups. In hypothesis 2, a bi-variate correlation was performed to see the relationship between experienced resistance with (a) ad-liking, (b) brand attitude and (c) referral intention,
respectively. Pearson’s r data analysis revealed a weak and insignificant positive correlation,
r(98) = .128, p = .204. The level of resistance to persuasion reported did not significantly relate
to ad-liking (Table 3). Therefore, hypothesis 2a was not supported.
Hypothesis 2b examined the level of experienced resistance (M = 3.05, SD = 1.16) and brand attitude (M = 5.63, SD = 1.37). Here, the Pearson’s r data analysis revealed a very weak
and insignificant positive correlation, r(98) = .05, p = .624. The level of resistance practitioners experienced had no significant correlation with their reports of brand attitude. Hypothesis 2b was also not supported.
Finally, amongst ninety-nine of the health care practitioners, the level of experienced resistance (M = 3.05, SD = 1.16) and referral intention (M = 5.01, SD = 1.32) was examined for hypothesis 2c. Again, the Pearson’s r data analysis revealed a very weak and insignificant positive correlation, r(97) = .014, p = .888. The level of resistance practitioners experienced had no significant correlation with their intention to refer the product displayed in the advertisement. As it is, 2c was not supported.
Hayes, 2018, Hayes’ PROCESS v3.0 was used to test Hypothesis 3 for the mediating effects of resistance (M) on the relationship between ego-involvement (IV) and (a) ad liking (b) brand attitude and (c) referral intention (DV’s; refer to Figure 11 for model). The indirect effect of ego-involvement on (a) ad liking (b) brand attitude and (c) referral intention through
resistance were not statistically significant; (a) indirect effect = -.02, SE= 0.03, CI=[-0.09, 0.02] (b) indirect effect = -.01, SE= .04, CI=[-0.11, 0.08] (c) indirect effect = .00, SE= 0.05, CI=[-0.12, 0.09]. The results showed (Table 4) that as a mediating variable, resistance did not statistically decrease the influence of ego-involvement on (a) ad liking (b) brand attitude and (c) referral intention. Thus, H3a H3b, and H3c were not supported. Therefore, involvement did not
significantly decrease resistance, which in turn did not affect ad-liking, brand attitudes or referral intentions (see Table 4).
For hypothesis 4, those exposed to a high ego-involvement message were proposed to have higher levels of experienced relevance as compared to those exposed to a low
relevance of M= 5.13 (SD= .97). By comparison, the low ego-involvement group (N= 49) was associated with a numerically lower level of personal relevance M= 4.97 (SD= .81). To test the hypothesis that the high ego-involvement and the low ego-involvement groups were associated with statistically significant different mean personal relevance levels, an independent samples t-test was performed. The outcome variable was found to be normally distributed (see Table 2) and equal variances are assumed based upon results of Levene’s test, F(98)= 2.50, p= .117. The independent samples t-test was not associated with a statistically significant effect, t(98)= -.908,
p= .366, two-tailed.
In hypothesis 5, the relationship between reported relevance with (a) ad-liking, (b) brand attitude and (c) referral intentions was examined. For hypothesis 5a, one hundred healthcare practitioners were surveyed about their level of experienced relevance (M = 5.05, SD = 0.89) and ad-liking (M = 3.58, SD = 0.78). A bi-variate correlation was conducted and a Pearson’s r data analysis revealed a moderate positive correlation, r(98) = .493, p<.001. Participants’ reported levels of higher relevance corresponded with higher ad-liking. The scatterplot in Figure 8 provides a visual of the positive correlation between relevance and ad-liking. Hypothesis 5b examined the level of experienced relevance (M = 5.05, SD = 0.89) and brand attitude (M = 5.63,
SD = 1.37). Again, the Pearson’s r data analysis revealed a moderate positive correlation, r(98) =
.510, p<.001. Practitioners who experienced more relevance, reported higher ad-liking. A scatterplot of the correlation between relevance and brand attitude will also show the positive relationship (see Figure 9). Finally, amongst ninety-nine of the health care practitioners (all who reported their intent to refer the product), the level of experienced relevance (M = 5.05, SD = 0.89) and referral intention (M = 5.01, SD = 1.32) was examined. Here, the Pearson’s r data analysis revealed a weak positive correlation, r(97) = .391, p<.001. Practitioners who
experienced more relevance, reported higher intention to refer the product displayed in the advertisement. Figure 10 shows a scatterplot of the positive but weak correlation between relevance and referral intention. Table 3 lists all variables in a correlation matrix with significance levels tagged.
Table 3
Correlation matrix of variables.
1. Ego-Involvement 2. Resistance 3. Relevance 4. Ad-Liking 5. Brand Attitude 6. Referral Intentions 1 -2 -0.12 -3 0.09 0.09 -4 0.02 0.13 0.49** -5 0.18 0.05 0.51** 0.67** -6 0.08 0.01 0.39** 0.66** 0.65**
Table 4
Effects of variables.
Pathway Estimate SE BC 95% CI
H1 a1 EGO —> RESIS -0.29 0.23 (-0.75, 0.18)
H2 b1.i RESIS —> AD 0.09 0.07 (-0.05, 0.22)
b1.ii RESIS —> BR.ATT 0.03 0.10 (-0.18, 0.23)
b1.iii RESIS —> REF.INT 0.16 0.12 (-0.21, 0.24)
H3/H6 c.i EGO —> AD 0.03 0.16 (-0.28, 0.34)
c.ii EGO —> BR.ATT 0.49 0.27 (-0.05,1.03)
c.iii EGO —> REF.INT 0.22 0.27 (-0.31, 0.75)
c.i’ EGO —> RESIS/REL —> AD -0.02 0.14 (-0.30, 0.26) c.ii’ EGO —> RESIS/REL —> BR.ATT 0.37 0.24 (-0.10, 0.85) c.iii’ EGO —> RESIS/REL —> REF.INT 0.15 0.25 (-0.35, 0.65)
H4 a2 EGO —> REL 0.16 0.18 (-0.19, 0.52)
H5 b2.i REL —> AD 0.43*** 0.08 (0.28, 0.59)
b2.ii REL —> BR.ATT 0.76*** 0.13 (0.49, 1.03)
b2.iii REL —> REF.INT 0.60*** 0.14 (0.31, 0.88)
Note 1: ***p < 0.001
Note 2: EGO — ego-involvement; RESIS — resistance; AD — ad liking; BR.ATT — brand attitude; REF.INT — referral intentions; SE — standard error; BC — bias corrected; CI — confidence interval.
Hayes’s PROCESS also tested Hypothesis 6 for the mediating effects of relevance on the relationship between ego-involvement and (a) ad liking (b) brand attitude and (c) referral
intention (refer to Figure 11 for model). The indirect effect of ego-involvement on (a) ad liking (b) brand attitude and (c) referral intention through relevance were not statistically significant; (a) indirect effect = .07, SE= 0.08, CI=[-0.08, 0.23] (b) indirect effect = .12, SE= .14, CI=[-0.14,
0.42] (c) indirect effect = .07, SE= 0.11, CI=[-0.13, 0.29]. The results showed (Table 4) that as a mediating variable, relevance did not statistically decrease the influence of ego-involvement on (a) ad liking (b) brand attitude and (c) referral intentions. Thus, H6a H6b, and H6c were not supported. involvement did not significantly decrease resistance, which in turn did not affect ad-liking, brand attitudes or referral intentions (see Table 4).
Discussion
The purpose of this study was to assess the influence of ego-involvement on ad-liking, brand attitudes and intentions to refer a product, with the role of resistance and relevance as potential mediating factors of ego-involvement. A closer examination of the differences in experienced resistance and reported relevance between high and low ego-involvement groups were also addressed. Participants in both groups had generally experienced lower levels of resistance and moderate levels of relevance. However, the results showed that there was little to no difference between the high and low ego-involvement conditions with regards to resistance or relevance levels. In effect, there was no significant relationship found between ego-involvement and resistance or between ego-involvement and relevance. This means that healthcare
practitioners’ level of resistance was not significantly lower when the advertisement was framed with higher levels of ego-involvement as compared to an advertisement that was framed with little/no ego-involvement. As there has been conflicting evidence of one’s vulnerability against resisting ego-induced messages (Halverson & Pallak, 1978; Pfau et al., 1997), it can be seen that there is no significant influence in the context of healthcare practitioners and health-related products. Also, reported relevance was similar amongst the two groups signifying that the framing of ego-involvement was not a determinant of relevant association with the product. This is inconsistent with previous findings which have addressed the positive relationship between
ego-involvement and relevance (Gendolla, Brinkmann, & Scheder, 2008). An explanation for this is that the main advertisement was created to be neutral and the ego-manipulation did not induce any greater attachment to the message as intended.
When observing the level of resistance and it’s effect on ad-liking, brand attitude or referral intentions, no significant results were found. Resistance was not found to be an important mediating factor between ego-involvement and ad-liking, brand attitudes, or referral intentions. However, many studies that have measured the influence of resistance on consumer behavior has shown ample evidence of it’s direct effect on consumer behavior (Gas, 2015) which may suggest that an error in this research design had caused the results to be insignificant.
Findings do support the notion that relevance is significantly correlated with ad-liking, brand attitudes, and referral intentions. This shows that the more relevant the advertisement is, the more likely brand attitudes, ad-liking and referral intentions will be higher than if the advertisement was irrelevant. This was also found in studies which focused on involvement (as defined as personal relevance) with an advertising message on brand attitudes (Cacioppo, Petty, Chuan, Rodriguez & Sarason, 1986)
Finally, the indirect effect of ego-involvement on ad-liking, brand attitudes, and referral intentions with a parallel mediation through resistance and relevance was assessed. Results proved to be insignificant in this case. This means that healthcare practitioners’ ad-liking, brand attitude, and referral intentions were not even indirectly influenced by level of ego-involvement through resistance or relevance. These findings are opposite of what has been found in other research pertaining to involvement and brand attitudes. Other studies have showed that high involvement with an advertising message enhances accessibility of brand attitudes (Kokkinaki, 1999).
Limitations and Future Research
Within the involvement construct, ego-involvement is better known as the type of involvement that is enduring. In terms of experimental manipulation, it can be thought of as a pre-existing condition and difficult to manipulate. Previous studies have had success with assessing the level of ego-involvement by means of manipulating attention or situational involvement, which is involvement induced by relation to the task at hand (Pfau, 1997). However, future research should first measure the level of ego-involvement then induce exposure to the advertisement and finally measure the corresponding variables in focus.
Limitations in the measurement of the ego-involvement construct are most likely the cause of the results being insignificant. Further research using better methods of measuring this construct is advised. The manipulation of ego-involvement in an experimental condition suggests that external validity is limited. Also, the sample cannot be generalized to all consumers.
Implications
Theoretical implications suggest that ego-involvement is a complex phenomenon to manipulate in an experimental setting, so it is difficult to establish true effects. Practical implications from this study lend to advertisers and marketers alike on the importance of relevance in positive attitude formations. This study has supported prior evidence of the effects of relevance on ad-liking and brand attitudes. It did contribute to the addition of referral
References
Ajzen, I., & Fishbein, M. (1980). Understanding attitudes and predicting social behavior. Englewood Cliffs, NJ: Prentice-Hall.
Cacioppo, J. T., Petty, R. E., Kao, C. F., & Rodriguez, R. (1986). Central and peripheral routes to persuasion: An individual difference perspective. Journal of personality and social
psychology, 51(5), 1032.
Fransen, M. L., Smit, E. G., & Verlegh, P. W. (2015). Strategies and motives for resistance to persuasion: an integrative framework. Frontiers in psychology, 6, 1201.
Gass, R. H., & Seiter, J. S. (2015). Persuasion: Social influence and compliance gaining. Routledge.
Gendolla, G. H., & Richter, M. (2006). Ego-involvement and the difficulty law of motivation: Effects on performance-related cardiovascular response. Personality and Social
Psychology Bulletin, 32(9), 1188-1203.
Gendolla, G. H., Brinkmann, K., & Scheder, D. (2008). Ego involvement moderates the assimilation effect of affective expectations. Motivation and Emotion, 32(3), 213-220. Halverson, R. R., & Pallak, M. S. (1978). Commitment, ego-involvement, and resistance to
attack. Journal of Experimental Social Psychology, 14(1), 1-12.
Huh, J., & Langteau, R. (2007). Presumed influence of direct-to-consumer (DTC) prescription drug advertising on patients: The physician's perspective. Journal of Advertising, 36(3), 151-172.
Jacks, J. Z., & Cameron, K. A. (2010). Strategies for resisting persuasion. Basic and
Jenkins, M., & Dragojevic, M. (2013). Explaining the process of resistance to persuasion: A politeness theory-based approach. Communication Research, 40(4), 559-590.
Klein, G. S., & Schoenfeld, N. (1941). The influence of ego-involvement on confidence. The
Journal of Abnormal and Social Psychology, 36(2), 249.
MacKenzie, S. B., Lutz, R. J., & Belch, G. E. (1986). The role of attitude toward the ad as a mediator of advertising effectiveness: A test of competing explanations. Journal of
marketing research, 130-143.
Madden, T. J., Allen, C. T., & Twible, J. L. (1988). Attitude toward the Ad: An Assessment of Diverse Measurement Indices under Different Processing" Sets". Journal of marketing
research, 242-252.
Muehling, D. D., & Laczniak, R. N. (1988). Advertising's immediate and delayed influence on brand attitudes: Considerations across message-involvement levels. Journal of
advertising, 17(4), 23-34.
O'Boyle, E. J., & Dawson, L. E. (1992). The American Marketing Association code of ethics: Instructions for marketers. Journal of Business Ethics, 11(12), 921-932.
Pfau, M., Tusing, K. J., Lee, W., Godbold, L. C., Koerner, A., Penaloza, L. J., ... & Yang, V. S. H. (1997). Nuances in inoculation: The role of inoculation approach, ego‐involvement, and
message processing disposition in resistance. Communication Quarterly, 45(4), 461-481. Riddick, F. A. (2003). The Code of Medical Ethics of the American Medical Association.
The Ochsner Journal, 5(2), 6–10.
Schminder, E., Ziegler, M., Danay, E., Beyer, L., & Bühner, M. (2010). Is it really robust? Reinvestigating the robustness of ANOVA against violations of the normal distribution.
European Research Journal of Methods for the Behavioral and Social Sciences, 6,
147-151.
Snyder, W. (2011). Making the case for enhanced advertising ethics: How a new way of thinking about advertising ethics may build consumer trust. Journal of Advertising Research,
51(3), 477-483.
van Reijmersdal, E. A., Fransen, M. L., van Noort, G., Opree, S. J., Vandeberg, L., Reusch, S., ... & Boerman, S. C. (2016). Effects of disclosing sponsored content in blogs: How the use of resistance strategies mediates effects on persuasion. American Behavioral Scientist,
60(12), 1458-1474.
Yagci, M. I., Biswas, A., & Dutta, S. (2009). Effects of comparative advertising format on consumer responses: The moderating effects of brand image and attribute relevance.
Journal of business research, 62(8), 768-774.
Yavas, U., & Riecken, G. (2001). A comparison of medical professionals with favorable and unfavorable attitudes toward advertising: An empirical study. Health marketing
quarterly, 18(3-4), 13-26.
Zaichkowsky, J. L. (1994). The personal involvement inventory: Reduction, revision, and application to advertising. Journal of advertising, 23(4), 59-70.
Zwier, S. (2017). “On the doctor's orders”: A pilot study of the effects of website marketing for medical specialist providers under gatekeeping arrangements. Health marketing
Appendix A
High/Low Ego-Involvement Manipulation *adapted from Laczniak (1989); Wright (1973)
Group Manipulation Instructions
High Ego-Involvement
Please look at the advertisement below and then answer the questions on the pages that follow.
ATTENTION: Although your participation will be kept anonymous, your response as an expert has been randomly selected as a direct influencer of price determination and subsequent promotion of this product. Please take extra care while viewing the ad,
We thank you for your professional feedback!
Low Ego-Involvement Please look at the advertisement below and then answer the questions on the pages that follow.
Appendix B
Manipulation Checks
Measure
Brand Awareness
…I am familiar with the brand. …I have use this brand before.
Endpoints: Strongly Disagree (1) / Strongly Agree (7)
Ego-Involvement
…Concerning the advertisement I just saw, I feel that my individual response will influence the
promotion of the product.
Endpoints: None at all (1) / A great deal (5)
Scales used for Measures
Measure
Resistance (by counter-argument)
*adopted from Fransen, Ter Hoeven, & Verlegh (2013); Zuwerink & Cameron (2003)
While reading, I… the information in the message. …contested
…refuted …doubted …countered
Endpoints: Strongly Disagree (1) / Strongly Agree (7)
Personal Relevance
*adopted from Zaichkowsky’s (1994) Personal Involvement Inventory
To me [OBJECT TO BE JUDGED] is: important _ _ _ _ _ _ _ unimportant* boring _ _ _ _ _ _ _ interesting relevant _ _ _ _ _ _ _ irrelevant* exciting _ _ _ _ _ _ _ unexciting*
means nothing _ _ _ _ _ _ _ means a lot to me appealing _ _ _ _ _ _ _ unappealing*
fascinating _ _ _ _ _ _ _ mundane* worthless _ _ _ _ _ _ _ valuable involving _ _ _ _ _ _ _ uninvolving* not needed _ _ _ _ _ _ _ needed *indicates item is reverse scored.
Ad Liking
*adopted from MacKenzie, Lutz, and Belch (1986) and Madden, Allen, and Twible (1988)
The advertisement was: …interesting
…good …likable …favorable …pleasant
Brand Attitude
*adopted from Ajzen and Fishbein (1980)’s 7-point bipolar scale
My attitude towards [BRAND] as a pain-reliever is: Bad _ _ _ _ _ _ _ Good
Unfavorable _ _ _ _ _ _ _ Favorable Negative _ _ _ _ _ _ _ Positive
Referral Intentions
…How likely are you to recommend [BRAND]’s pain-reliever to someone else? …How likely are you to refer [BRAND]’s pain-reliever to a patient?
Endpoints: Extremely Unlikely (1) / Extremely Likely (7)
