Message Framing

Message Framing

When to emphasize the negative in an online

retailing environment?

Jeroen Hoeve

Message Framing

When to emphasize the negative in an online retailing

environment?

Key Words: Message Framing, Purchase intention, Involvement,

Regulatory Focus Theory

Master Thesis

MSc Business Administration, Marketing Management

Author: Jeroen Hoeve

Address: Westersingel 2a, 9718 CK Groningen Phone Number: +31 (0) 6 397 995 31 Email: hoeve.jeroen@gmail.com

Student Number: 1975838

Organization: University of Groningen Faculty: Faculty of Economics and Business 1st _{Supervisor: Dr. Jia Liu}

2nd _{Supervisor: Stefanie Salmon}

Company: Kadaster Supervisor: Gaby Anink

Management Summary

In this research the effectiveness of message framing, and the moderating influence of involvement and regulatory focus theory, has been investigated. The main goal of the research is to get insight in the effectiveness of message framing in this specific context, since there have been contradicting results over the past few decades. Secondly, most research has focused on health behavior, so this research broadens the field for this topic as it takes place in the context of the real estate industry. More specifically, this research involves information products provided by Kadaster through one of their online channels. This lead me to the following research question:

“What is the influence of positive and negative message framing in a

product offering on online purchase intention?”

The research results confirm that the use of negative and positive message framing really is a strategy that has to be handled with care. Apposed to what I found in literature, I found that the interaction of involvement and negative message framing negatively impacts online purchase intention. Prevention focus and involvement have a significantly positive effect.

Apposed to what I proposed, I found that regulatory focus theory does not significantly impact purchase intention. One plausible reason for this finding might be that respondents did not recognize the opportunity of changing their effective state by purchasing the product. Another reason for this result is that some consumer goals fit regulatory focus theory better than the others. This also might be the case for consumers browsing the web for a new home.

Preface

During the marketing management program I have always been fascinated by reactions of consumers to different communication strategies. It is not just about getting to know characteristics of consumers but also how they behave in the market place. It gets even more interesting when you can find ways to influence consumers using cognitive processes. This thesis has given me the opportunity to combine the research field of consumer behavior while simultaneously getting insight into aspects relating to online retailing.

This thesis represents the final chapters of my life as a student. The process of conducting academic research has taught me a lot about myself. Once again, it has become clear that I really am a team player. Working by myself was not my favorite part of the deal. Therefore, I would like to thank all of my family and friends for supporting, inspiring and encouraging me. Especially my fellow students Annemieke, Annerieke, Natasja and Rob have been very supportive during the process, as they were “fellow sufferers”. Special thanks go out to my parents, who provided me with the opportunity to accomplish my academic objectives.

This thesis has taught me to work independently, manage my own time and develop my analytical skills. You can only learn by doing. I would like to thank Gaby Anink for her managerial support at Kadaster. Last but not least, I would like to thank my supervisor Dr. Jia Liu for her constructive feedback during the process. Last but not least, I thank Stefanie Salmon for helping with the final touches on my analyses and her general feedback.

Groningen, May 9 2012

1

Introduction

1.1

Background and Context

A question of considerable importance to a marketer offering a product is whether to frame the message to current or potential customers in terms of gains they can obtain from using the product, or in terms of losses they would suffer from not using it (Ganzach and Karsahi 1995). An example of both framing options is given below, related to the risks involved with smoking:

Positive: “You will reduce your risk of developing lung cancer should you quit smoking”

Negative: “You will increase your risk of developing lung cancer should you not quit smoking”

Over the past few decades a vast amount of research has been published stressing the (in)effectiveness of message framing (Block and Keller, 1995; Ganzach et al. 1995; Maheswaran and Meyers-Levy, 1990; Rothman and Salovey, 1997). In this previous research the effectiveness was measured using attitude judgements, behavioral measures, cognitive elaboration and frame related thoughts as dependent measures. One robust finding from literature that has emerged is that the relative effectiveness of negative versus positive message framing depends on the extent of cognitive elaboration the target audience engages in (Shiv, Edell Britton and Payne 2004). By cognitive elaboration, as mentioned in previous research, they mean the level of processing motivation and processing opportunity, which could lead to systematically processing the object. This research suggests that negative message framing is more effective when processing motivation is high.

The context in which this research takes place enables me to test the effectiveness of both methods for offering products, negative and positive message framing. The Dutch Land Registry Office, hereafter referred to as Kadaster, offers some of her products through Funda, an online channel for real estate companies, to communicate their offerings. Kadaster is a self-administering state body; this means that it is a legal entity under public law, which performs its tasks as an independent organization. Registering, informing, maintaining records and usage planning have been the tasks of Kadaster since 1832. Since Kadaster's tasks are mostly focused to help public services, it is challenging for the company to commercialize its products towards consumers.

Kadaster is communicated to the public as being a reliable independent reference point concerning real estate registration and spatial planning. Some products of Kadaster are offered through the website of Funda. In this way Kadaster can increase brand awareness, since this is a channel that reaches out to prospective homebuyers. Conversion through this channel is relatively low at the moment, so for this reason Kadaster is looking for possibilities to increase it. E-commerce managers in general are interested in what influences conversion, and how they can improve it by adapting the web site to consumer preferences (Montgomery, Li, Srinivasan, and Lichy, 2002). Evidence from online sales continues to show low conversion ratios, from enquiry to sales, in many online buying situations (Grant 2007; Montgomery et al. 2002). Companies flocked to the new way of doing business expecting to boost sales and increase profitability by adding customer value through lowering information search and acquisition costs (Klein 1998). Though, the key question is not whether to deploy Internet technology – companies have no choice if they want to stay competitive – but how to deploy it. A creative way to present persuasive messages is a tool that advertisers can use to break through clutter and stand out (Yoo 2005). Kadaster has no insight in what percentage of the total conversion originates from the web page on Funda. This makes it difficult to analyze any differences in conversion; therefore, purchase intention will be used in the context of this study.

interesting strategy to increase purchase intention, since it is subtle, and it does not involve a new marketing campaign. Furthermore, message framing is a technique Kadaster can use to influence buying behavior since they have control over this medium, and costs for applying such a strategy are relatively low.

The effectiveness of message framing has proven to be dependent on cognitive elaboration. For this reason we investigate the moderating effect of prevention and promotion focus (Mourali, Böckenholt and Laroche 2004; Greifender, Bless, and Pham 2011). Through this theory I will obtain a more clear view on the impact of the level of prevention and promotion focus on online purchase intention. This information will provide marketing managers with the opportunity to offer products according to consumer goals.

1.2

Research Questions

The main research question is formulated as follows:

“What is the influence of positive and negative message framing in an product offering on online purchase intention?”

1.2.1 Research Objective

The objective of this research is to examine how positive and negative message framing influences online purchase intention, by manipulating the website. I will also investigate which consumer characteristics moderate this relationship. In this context I will study the influence of involvement and regulatory focus theory. The latter meaning the influence of consumer’s promotion and/or prevention focus.

1.2.2 Research Questions

To answer the problem statement, the following research questions have bee formulated:

1. What is the influence of negative and positive message framing on online purchase intention?

2. How does prevention and promotion focus moderate the effect of message framing on online purchase intention?

1.2.3 Theoretical and managerial relevance

This study adds to the existing framing literature by examining how consumer characteristics moderate the impact of the framing strategy on purchase intention. It takes into account the situational differences of consumers in order to see what framing strategy is the most profitable.

Many studies have researched the effect of message framing in the health care industry (Rothman 2006; Meyers-Levy 1990; Wheatly 1970), it would be interesting to see how these theories apply to the real estate industry, and what role involvement plays in this relation. This research concluded that negative message framing is more effective than positive framing when the subject’s elaboration is high. Though, the dependent variables focused on attitude measures and change in health behavior. I, on the other hand, will investigate online purchase intention as the dependent variable. Financial risks refer to the risk of losing large sums of money, or spending too much on an inferior product (Fennis and Stroebe, 2010). In this case it is not so much about spending too much money on one of Kadaster's products but preventing to spend too much money on an inferior house when not using the products offered by Kadaster through Funda.

Another reason why this research contributes to current work is because the context, in which it takes place, is different from most of the previous studies. The message framing is used in order to change online purchase behavior, as opposed to changing behavior concerning health related topics.

What this study adds to the existing knowledge and research is that I do not only show that the effectiveness of the use of message framing is context dependent, but also investigate the moderating effect of prevention and promotion focus (Mourali, Böckenholt and Laroche 2004; Greifender, Bless, and Pham 2011). These focuses represent the two broad categories in which consumer goals can be classified. This information will provide marketing managers with the opportunity to offer products according to consumer goals. Previous research has mainly focused on the role of involvement, which has proven to be relevant, but I do want to see if there are any other factors that influence the effectiveness .

industry in general, should consciously think about how to approach prospective buyers/sellers. Maheswaran and Meyers-Levy (1990) state that it seems worthwhile to explore whether, and under what conditions, the framing outcomes will hold for non-health-related, more conventional products, issues, and behaviors that are likely to entail less risk or uncertainty than preventive health behavior .

Using an online questionnaire, the relative strength of message framing, for influencing consumer behavior in an online environment, will be tested. This will give valuable information with respect to the effect of message framing, and if it is worthwhile for Kadaster to use this technique in order to increase firm performance.

1.3

Structure of this Thesis

2

Theoretical Framework

This chapter will deal with the theoretical framework that will, ultimately, be the basis for the conceptual framework. The conceptual model is presented at the end of this chapter. Each variable included in the conceptual model will be presented in a separate paragraph. At the end of each paragraph the hypotheses will be presented.

2.1

Literature Review

2.1.1 Message Framing

Meyers-Levy et al. (1990) introduce positive and negative framing as presenting the equivalent information in two different ways. Messages that are framed positively stress either the benefits gained or the negative consequences avoided if one accepts a course of action. Negatively framed messages stress either the negative consequences incurred or the benefits foregone if one does not accept such action (Lee and Aaker, 2004).

The theoretical base for this research is derived from the Prospect Theory (Smith and Petty 1996; Maheswaran et al. 1990). It suggests that people are risk-averse when a decision problem is formulated in terms of loss and risk-prone when the problem is framed in terms of gain. Second, it suggests that people exhibit loss-aversion, i.e. that losses loom larger than gains.

Negatively framed information might also arouse more fear than messages that focus on gains from engaging in behavior and thus, might be more persuasive for this reason. Several studies have found no difference between negatively and positively framed messages on their measures of fear arousal, suggesting that the fear arousal interpretation was not plausible. The results of the research conducted by Smith and Petty (1996) suggest that employing message framing that is unexpected can increase the extent to which messages are elaborated.

when their expectancies were confirmed but increased their level of processing when their expectancies about the message were disconfirmed. This is in line with other investigations that have attempted to create conditions of expectancy confirmation and disconfirmation.

Are positive frames always more persuasive than negative frames for communications? Some research suggests that when subjects carefully process information negative framing will be more persuasive, because negative information is regarded to be more informative than comparable positive information.

Meyerowitz and Chaiken (1987) indicate that with a lower level of efficacy, that is, when it is uncertain that the recommendations will lead to the desired outcome, negative frames are more effective than positive frames.

Ganzach and Karsahi (1995) examine the effect of framing on the persuasiveness of a message in the financial domain, rather than in the health domain.They found that negative framing has a much stronger effect on the behavior of credit card owners than positive f raming. The research implicated that the framing also resulted in behavioral change. Utilization was more than double when the message was framed in terms of loss than when it was framed in terms of gain, and the charges made by customers who received the loss-framed and used the card were also more than twice those made by customers who received the gain-framed message. Regarding word-of-mouth, Herr (1991) provides direct support for the proposition that negative information is more informative and has a bigger influence on consumers’ decision making than positive framing. This is in line with what the research Block et al. (1995) suggests.

Research has also documented the effect of framing in the cognitive and attitudinal domains. Loss framing resulted in both better recall and stronger persuasiveness of the message, as well as higher involvement with the payment method (Ganzach and Karsahi, 1995). In this context the attitudinal effect on involvement is particularly interesting. This result is relevant for marketers, because involvement may be an important factor in inducing long-term behavioral change.

turn out to be more risk-averse because of the negative framing condition they are more likely to purchase the product. Therefore, it can be concluded that message framing could influence purchase intention.

All in all, it is difficult to decide what framing strategy yields the best results. It appears to depend on the specific environment in which the framing takes place. Since there have been opposing results in past literature concerning message framing and the effectiveness thereof I decided that I will not hypothesize that there is a positive or negative relation between the independent (message framing) and dependent variable (conversion). Past literature has shown that it depends on the factors that could be moderating the relationship. The moderators, relevant for this research, will be presented in the next paragraphs. Since the effectiveness of negative and/or positive message framing depends on the specific situation I will investigate which framing strategy would be most effective for Kadaster.

The conceptual model will not only clarify the proposed relations but also emphasizes that message framing might have an effect on buying behavior in this context. But it depends on the specific characteristics of the website visitor.

2.1.2 Involvement

Involvement has been defined as an individual, internal state of arousal with intensity, direction, and persistence properties (Zhang et al. 1999). Research has defined it using learning situations, high-involvement learning situations and low-involvement learning situations. A high low-involvement learning situation is one in which the consumer is willing and motivated to process or learn the material. On the contrary, low-involvement learning is when the consumer is not willing or has no motivation to learn the material (Hawkins, Hoch and Meyers-Levy, 2001).

cues are the non-essential elements in a communication that are not inherently related to the central merit of the attitude object. The different ways in which the same ad message is conveyed to consumers are expected to function as peripheral cues.

Meyerowitz and Chaiken (1987) investigate the role of message framing as a tool to induce behavior in relation to breast self-examination. In this research, the level of perceived efficacy was used as a moderator. Perceived efficacy meaning the level of certainty that the recommendations lead to the desired outcome. This research suggests that negative framing is more persuasive when self-efficacy is high. Ganzach and Karsahi (1995) concur with these results. Block and Keller (1995) suggest that uncertainty, in relation to following the recommendations, causes more in-depth processing, in which case negative message framing is more persuasive. In the same vein, when there is certainty about following the recommendation, less effortful processing will occur. For this scenario, positive message framing is more persuasive. This is consistent with Maheswaran et al. (1990), which, in the same line of reasoning, found that negative framing is more persuasive when in-depth processing occurs, or as how they state it; when involvement is high. I will elaborate more extensively on involvement in the next paragraph.

Taking this information into account, I can summarize this as involvement being the level of willingness and motivation with which the subject processes the object in question. According to the heuristic-systematic model (HSM) theory, when people rely predominantly on systematic processing, due to the high personal relevance of the message, negatively versus positively framed messages should be more persuasive. Those who employ systematic processing base their judgments on detailed scrutiny and the perceived diagnostic of message data (Meyers-Levy et al. 2004).

The framing suggestion of prospect theory holds that people evaluate information regarding uncertain (risky) alternatives in terms of either potential gains (positive framing) or potential losses (negative framing) and that preference can be altered by the way information is presented. In brief, negative message framing is suggested to make risky options seem more desirable (Smith and Petty 1996).

Negatively framed information might receive different weight in participants’ judgments than positively framed information. This would be consistent with several studies in impression formations that have shown negative trait information to exert a disproportionate influence on personality judgments relative to positive trait information.

Two types of involvement have been discussed frequently in this research field: situational and enduring involvement. Search engine users are regarded as consumers with high situational involvement. Consumers’ level of involvement has been found to influence the effect of message framing. Consumers with high involvement pay more attention to the issue of losses instead of gains, and have a stronger tendency to show lower sensitivity to any type of gains, concentrating on minimizing losses. On the contrary, consumers with low involvement pay more attention to the issues of gains, concentrating on maximizing any type of gains. Hence, negative framing is more effective than positive framing when involvement is high, whereas the opposite is true when involvement is low (Yoo, 2005).

Chaiken (1994) implies, applied to persuasion systematic processing, that people have formed or updated their attitudes by actively attending to and cognitively elaborating persuasive argumentation. In contrast, heuristic processing implies that people have formed or changed their attitudes by invoking heuristics such as “experts can be trusted”, “majority opinion is correct”, and “long messages are valid messages”. These dual-process theories regard systematic processing as more effortful and capacity limited than heuristic processing. Heuristic processing predominates when motivation or capacity for effortful processing is low.

communication is likely to stand out and processed more carefully by the subjects. Consumers in the low involvement condition should be more appealed to positive message framing as they pay more attention to the potential gains.

This theory leads to the hypothesis that negatively framed messages should be more persuasive than positively framed ones when issue involvement is high (Loken 2006), but the reverse outcome should emerge when issue involvement is low (Maheswaran et al. 1990).

H1: When consumers’ involvement is high, negative message framing has a stronger positive effect on online purchase intention than positive message framing.

The results of this hypothesis should help Kadaster on deciding how to communicate their product offering through this channel. When involvement, generally speaking, is high under website visitors Kadaster can adjust its product offering on the website according to the implications of the literature. This is also applicable for the opposite results.

2.1.4 Regulatory Focus Theory

With the increasing recognition, in recent years, of the importance of goals and motivation in shaping consumer behavior, researchers have found, in regulatory focus theory, a powerful and parsimonious framework for investigating various phenomena such as persuasion and choice decisions (Mourali, Böckenholt and Laroche 2004).

Both promotion and prevention focus systems are assumed to exist in each person and can be activated separately depending on the situational demand. Mourali et al. (2004) state that prevention-focused consumers are expected to display a greater susceptibility to the compromise effect, whereas promotion-focused consumers would be more susceptible to the attraction effect. Because promotion focused consumers use an eager strategy for achieving hits and ensuring advancement they should be more sensitive to the dominance heuristic. Heuristic modes of evaluation occur under promotion versus more systematic processing under prevention and indicating a preference for speed versus accuracy in task completion under promotion focus.

If this were actually true, it would mean that, for prevention focused consumers, they would be more attracted by negatively framed messages, since it stresses the negative consequences incurred or the benefits foregone. Because negative feelings are unpleasant people are motivated to reduce, or even eliminate, these feelings. Consequently, if people perceive that an activity can accomplish this, they may be more attracted to it than they otherwise would be. When individuals recognize that the behavior can potentially reduce or eliminate the negative feelings they are experiencing, motivational influences predominate (Shen and Wyer 2008).

One study suggests that that priming ideals (a promotion focus) encourages the reliance on affective feelings in judgment, whereas priming ought’s (prevention focus) discourages it (Pham and Avnet 2004). Additional research showed that this is because feelings are perceived to be more informative under a promotion focus than a prevention focus. (Andrade and Cohen 2004) suggest that motivational influences of negative affect depend on whether participants recognize the affect-regulating capabilities of a behavior at the time they consider performing it. Meaning that one will only be motivated to change behavior when on recognizes that the behavior can change the affective state.

The recognition is only likely to occur if the negative feelings one is experiencing and their antecedents are conceptualized at a level of specificity that makes salient the source of one’s negative affect and consequently the activity of behavior that reduces it. Not all evaluations are based on feelings. Utilitarian criteria concern the extent to which an object or an event satisfies people’s values, goals, or needs, whereas hedonic criteria concern the feelings that the stimulus elicits. People typically only use their feelings as a basis for evaluations that are based on hedonic rather than utilitarian considerations.

Since the context of this study will modify the website content, and specifically the text in relation to the product offering, it seems like these elements will affect the utilitarian elements of the object. Since feelings are suggested to only be used as a basis for evaluations based on hedonic considerations it follows that feelings will not affect the evaluation of the product offering on the Kadaster web pages.

Furthermore, lower efficacy motivates more in-depth processing, whereas higher efficacy evokes less effortful processing. This demonstrates that processing effort is a significant factor in the process underlying the framing effect (Block et al. 1995)

Hypotheses

In order to get a clear view on what this means for the context in which this research takes place I will elaborate on this by introducing the hypotheses related to promotion and prevention focus.

Therefore I suggest that, in comparison to negative framing, I expect promotion focus to positively moderate the influence of positive message framing on conversion.

H2a: When consumers’ promotion focus is predominant, positive message framing has a stronger effect on online purchase intention than negative framing.

Negative message framing on the other hand will be more appealing to prevention-focused consumers. Therefore I suggest that, in comparison to positive message framing, I expect prevention focus to positively moderate the influence of negative message framing on conversion.

2.2

Conceptual Model

Using the previous overview on current literature, a conceptual model has been created below in order to clarify the research hypotheses. The advertising characteristics include both of the message framing conditions. The consumer characteristics consist of promotion and prevention focus, based on regulatory focus theory, and involvement.

Conceptual model

3

Methodology

This chapter represents an overview of the research method that will be used to test the hypotheses that were formulated in the second chapter. I start by defining the type of research and which metrics will be used. Ultimately I will conclude by presenting the plan of analysis.

3.1

Type of Research

In order to analyze the research question ‘What is the influence of positive and negative message framing in an advertisement on online purchase intention?’ a 2 (positive and negative framing) x 2 (high involvement and low involvement) x 2 (promotion focus and prevention focus) between subjects factorial design was implemented. This research deals with primary data, since I collect the data first hand, for this specific research. Research can be classified as exploratory or conclusive. Exploratory research has the primary objective of providing insights into, and an understanding of, the problem confronting the researcher. Generally taken, it is followed by further exploratory or conclusive research. Conclusive research, on the other hand, has the main purpose of testing specific hypotheses and examines relationships. The findings are used as input for decision-making. Two distinctions within conclusive research can be made: descriptive and causal research. Descriptive research focuses on a specific situation. Causal research pursues the goal of finding a cause-effect. Considering the above this research is qualified as being causal (Malhotra 2007). The purpose of this research is to examine what the effect is of moderating influences on the relationship between message framing and conversion.

3.2

Procedure

A questionnaire will be used in order to collect the data. This questionnaire will be made available online, on the specific website where Kadaster is offering their products to consumers. The measurement items used in this questionnaire are dealt with in the subsequent paragraphs. A banner will be placed on the website in which I encourage website visitors to complete the questionnaire. Participants have the opportunity to leave their email address in order to potentially win a VVV-Voucher, valued at 50 Euros.

assigned to either the negative message framing condition or the positive framing condition. The image of the website will be shown with one of the framing conditions. After this I will measure respondent’s purchase intention. The questionnaire itself is included in appendix I.

3.2.1 Independent variable

The independent variable in this research is message framing. Messages that are framed positively stress either the benefits gained or the negative consequences avoided if one accepts a course of action. Negatively framed messages stress either the negative consequences incurred or the benefits foregone if one does not accept such action (Meyers-Levy et al. 1990). The product offering on the web pages of Kadaster will be manipulated accordingly (appendix II).

3.2.2 Dependent variable

Online purchase intention is a ratio variable through which people indicate the likelihood of purchasing one of Kadaster’s products. I measured this variable through four items, all 7-point Likert scales, according to Bart, Shankar, Sultan, and Urban (2005). The item “I would register at this site” was left out since it was not applicable for the context of this study. The items that were used can be found in appendix I.

3.2.3 Moderating variables

Involvement

Involvement is measured according to Bearden, Haws and Netemeyer (2001). The components of involvement are measured using 16 Likert-type statements (strongly agree to strongly disagree), all on 7-point scales (Appendix III). These items represent five dimensions of involvement to form an overall profile. The five dimensions represented are:

- The perceived importance and risk of the product class - The subjective probability of making a miss purchase - The symbolic sign or value attributed

Regulatory focus

According to Higgins et al. (2001) the following items are used, some adjusted for this particular research. This set of questions asks participants about specific events in their life (Appendix III). A 7-point Likert-scale is used, 1 indicating this event took place never or seldom, and 7 indicating this event occurred very often. Items 1, 3, 7, 9, 10, and 11 are Promotion scale items. Items 2, 4, 5, 6, and 8 are Prevention scale items. Questions 1,2,4,6,8,9,11 are reversed scaled.

3.2.4 Manipulation check

In order to check the manipulation of the message framing I will include an item in order to check the valence of the product offering, according to Khare, Labrecque and Asare (2011). The manipulation check consists out of three questions, all 7-point Likert scales (Appendix I). The first question focuses on the valence of the message, i.e. “I would say the offering of the products by Kadaster is positive/negative”. The following two questions will test whether respondents think the offering emphasizes benefits foregone or risks they will be exposed to when not buying the product, i.e. ‘ The product offering on this web page emphasizes the negative consequences you prevent when you buy an information product’ and ‘The product offering on this website emphasizes the risk you are exposed to when not buying an information product’. This will test if respondents actually perceive the message to be positively or negatively framed.

3.3 Data Collection

This research uses a survey technique by using an online questionnaire, using an online tool called MWM2. Several types of surveys can be designed but this research will focus on personal surveys. The advantage of an online survey is that it is easily accessible for respondents in order for me to get the desired response rate. In addition, it will assure me with the ability of comparing results. On top of that, it increases speed and accuracy and will facilitate data processing (Malhotra, 2010).

agreement or disagreement with each of a series of statements about the stimulus objects. The interval scale ranges from strongly agree to strongly disagree on a 7-point scale. The survey also includes some open-ended questions for the control variable ‘Age’.

3.4

Population and Sample

The sampling design process will consist out of four steps; defining the target population, determining the sampling frame, selection of the sampling technique, and determining the sample size (Malhotra, 2010).

The sampling design begins by defining the target population; this is the collection of elements or objects that possesses the information needed for this research (Malhotra, 2010). The target population for this research are prospect homebuyers and home sellers using the Internet, and in particular www.funda.nl for their orientation in the real estate market. The next step is determining the sampling frame, which is a representation of the target population. It consists of a list or set of directions for identifying the target population (Malhotra, 2010). This research will be conducted using an online survey tool, which will be made available through a banner. The survey is available for all Internet users accessing this particular website.

This study will use a non probability sampling technique. What characterizes this technique is that the respondent happens to be at the right place at the right time. The reason why this technique fits with this research is because the respondents happened to be visiting the website and willing to fill out the questionnaire.

The final stage of this process is determining the sample size. This refers to the number of elements that will be included in this research. Since this research covers a 2 (negative framing, positive framing) x 2 (high involvement and low involvement) x 2 (promotion focus and prevention focus) between participants’ design a minimum of 160 participants is necessary, when accounting for 20 participants per measurement item.

3.5

Plan of Analysis

3.5.1 Reliability

Reliability refers to the extent to which a scale produces consistent results if repeated measurements are made. We will check the reliability of the concepts using the Cronbach’s alpha. In table 2 an overview is presented, which shows the Cronbach’s Alpha per measurements. The Cronbach’s alpha is the average of all possible split-half coefficients resulting from different ways of splitting the scale items. This coefficient varies from 1 to 6, and a value of 0.6 or less generally indicates unsatisfactorily internal consistency reliability. Item 16 (“Real estate is a topic which leaves me totally indifferent”) of the involvement measure was deleted for further analysis (appendix III), in order to achieve a satisfactory internal consistency.

3.5.3 Multiple Regression

4!

Results

4.1

Descriptives

A total of 165 (N=165) male and female participants (84 male, 81 female; Mean age= 41.14 years (range 18-73), SD= 17.520) participated in this research.

Gender & Age

Gender Mean Age Range

84 Male (50.9%) 81 Female (49.1%)

41.14 55 (18-73)

Table 1: Demographic profile

Between Groups Descriptives; Gender & Age

Table 2: Between Groups Descriptives

The mean age in the positive framing condition was 44.7 and 35.8 in the negative framing condition, respectively. In the positive framing condition 58,1% of the participants was male and 41,9 percent was female. In the negative framing condition 43% was male and 57% female. An independent Sample T-test shows that there is a significant (P<0.001) difference between the mean age of the two framing conditions (Appendix VII).

0

15

30

45

60

Mean Age

% Men

% Women

t

Sig. (two-tailed)

Age

3,821

0.000***

# # Table&3:&summary&(*=90%&con7idence&level,&&**=95%&Con7idence&level,&***=99%&& & & con7idence&level)

Another independent samples T-test shows that there is a moderately significant (P<0.100) difference between the framing conditions in gender (Appendix 8).

t

Sig. (two-tailed)

Gender

-1.949

0.053*

& & & Table&4:&summary&(*=90%&con7idence&level,&&**=95%&Con7idence&level,&***=99%&& & & & con7idence&level)

The respondents were from all walks of life, as is resembled by their education levels shown in table 5.

Between Groups Descriptives; Education

Table 5: Education level (Dutch standards)

Positive Framing

Negative Framing

4.2

Manipulation Check

In order to check if respondents perceived the valence of the message framing differently, a oneway ANOVA (analysis of variance) is performed. This test will give me insight in the variance between the two framing conditions concerning valence of the message. First of al the three items of the construct manipulation check are computed in order to obtain the variable Manipulation_Check. The F value is 4,381 and the ANOVA is significant (P<0,038**). Meaning that a significant level of the respondent rated the valence of the framing conditions differently between the subject groups, i.e. subjects actually experienced the negative framing condition to have a negative valence and vice versa.

Manipulation Check

Mean

Standard Deviation

F

Sig.

Positive

framing

condition

4.2558

1.44229

4,381

0.038**

Negative

framing

condition

4.7089

1.32837

&& & Table&6:&Results&ANOVA

4.4

Key Results

In order to analyze the interaction effects of involvement on online purchase intention I will estimate a multiple regression equation (n=165). The dependent variable is ‘Purchase Intention’ and the independent variables are ‘Involvement’, ‘Promotion Focus’, ‘Prevention Focus’, and ‘framing condition’ and in order to test the interaction all independent variables are multiplied by ‘Message Framing’, which had to be recoded using dummy variables (0= Positive Framing Condition; 1= Negative Framing Condition). Additionally, ‘Budget’, ‘Age’ and ‘Gender’ are included as control variables (Figure 2).

Linear Regression Equation

& Figure&2:&Linear&Regression&Equation

The overall regression equation is statistically significant (F=2.131; p<0.1), as is shown in table 3. The adjusted Rsquare, indicating the goodness of fit adjusted for the number of variables, is 0.225. This indicates that when all concepts are used as predictors, only 22,5% of the variance is explained by the regression model and 72,5% is residual variability.

Summary Linear Regression

Dependent

Variable

Adjusted R

square

F

Sig.

Purchase

intention

0.225

2.131

0.055*

Except for ‘Involvement’, ‘Prevention Focus, and the interaction effect of involvement and message framing none of the variables are significant, i.e. Purchase intention is impacted by ‘Involvment’, ‘Prevention and the interaction effect between message framing and involvement.

Therefore, values below 0.1 indicate issues relating to multicollinearity. Looking at this model, it can be concluded that multicollinearity is not a concern

Coefficients

Factor

β

t

Sig.

Tolerance

VIF

Involvement 0.627 1.956 0.060* 0.120 1.493 Promotion Focus -0.298 -0.769 0.448 0.368 2.715 Prevention Focus 0.333 1.740 0.093* 0.418 2.393 Message Framing -0.057 -0.143 0.887 0.882 1.134 Message Framing*Prevention Focus -0.209 -0.678 0.503 0.444 2.252 Message Framing*Promotion Focus 0.467 0.850 0.402 0.375 2.667 Message Framing*Involvement -2.058 -2.508 0.18** 0.728 1.374 Budget -0.041 -0.185 0.855 0.832 1.202 Age 0.030 1.672 0.105 0.636 1.573 Gender 0.360 0.825 0.416 0.679 1.472 Table&8:&&Summary&of&multiple&regression&results,&dependent&concept&=&‘Purchase&Intention’&(*=90%& con7idence&level&&&&&**=95%&Con7idence&level&&&&&***=99%&con7idence&level)

of the subject they yield better results in terms of online purchase intention. Prevention focus also has a significantly positive effect on purchase intention. As can be seen by the direction of the coefficient, purchase intention will increase by 0.333 if prevention increases by 1.

Additionally, looking at the direction and the value of Message Framing*Involvement it can be concluded that the interaction effect between these two variables has a significant (p<0,05) negative impact on online purchase intention, i.e. when the variable Message Framing*Involvement would increase with 1, the online purchase intention score would decrease by about 2.058. Therefore, H1 is not supported. This means that the combination of negative message framing and high involvement negatively impacts online purchase intention. This is not in line with what I expected. This could be explained by the fact that involvement was not high enough in order for subjects to analyze the product offering more carefully. This way consumers will realize that the product offered will be able to change their affective state after it being pointed out that there are risks involved.

Hypotheses

The multiple regression tests of the hypotheses were significant. Table 5 will give an overview whether the hypotheses were accepted. In the next chapter I will discuss the results and conclusions will be drawn. Subsequently, I will set up some managerial implications and discuss the limitations of this research.

Hypotheses

Hypothesis

Sig.

Supported?

H1:

negative message framing has a stronger effect on online purchase intention than positive message framing.

0.018**

No

H2a:

predominant, positive message framing has a stronger effect on online purchase intention than negative framing.

0.402

No

H2b:

predominant, negative message framing has a stronger effect on online purchase intention than positive framing.

0.503

No

5!

Summary

5.1 Conclusion

The main purpose of this study was to answer the following question: “What is the influence of positive and negative message framing in an advertisement on online purchase intention?” As can be concluded from chapter 4, involvement and prevention focus positively impacts online purchase intention.

Impact of message framing on online purchase intention

In the literature review on message framing I concluded that it depended on the specific environment which framing strategy would yield the best results for Kadaster. I can now confirm that the interaction of negative message framing and involvement impacts online purchase intention negatively, apposed to what I hypothesized. This means that the likelihood of someone purchasing an information product of Kadaster is smaller in the negative framing condition compared to the positive framing condition. Looking back at the framing literature Block an Keller (1995) implied that positive message framing would be more persuasive when there is certainty about following the recommendation and less effortful processing occurs. This is also in line with Maheswaran et al. (1990), who found that negative framing is more persuasive when in-depth processing occurs. In this case this would mean that the level of processing is not high enough in order to make negative framing a persuasive communication strategy.

Interaction effect of message framing and Involvement

This study confirms that involvement has a negatively moderating influence on negative message framing, compared to positive message framing, on purchase intention. This means that the negative framing condition is more persuasive for the website visitors who are highly involved with this particular issue. This disconfirms the theory of Meyers-Levy et al. (2004) that the degree of risky implications with an advocacy and its level of personal relevance can influence the type of processing that people employ.

Interaction effect of message framing and regulatory focus

purchase intention. Neither does the interaction effect of prevention focus and negative message framing impact online purchase intention. Lee et al. (2004) stressed that although any specific goal may be pursued with either a promotion or prevention focus, some goals are more compatible with a particular self- regulatory strategy, resulting in a higher level of fit. Looking back at this study, the goals primed in the manipulation might not have been compatible with either of the self-regulatory focus strategies. Andrade et al. (2004) suggest that people will only be motivated to change behavior when one recognizes that the behavior can change the effective state. In this case, it would be possible that participants did not recognize that the suggested behavior, of buying an information product, would change their effective state.

5.2 Managerial Implications

One of the main outcomes of this research is that the interaction of involvement and negative message framing negatively impacts online purchase intention. Because of this result, marketing managers at Kadaster should not look into possibilities of using negative message framing as a tool to induce buying behavior, whenever consumers’ involvement is likely to be high. Some channels, through which the products are offered, might attract high involvement customers. Negative message framing does not help in increasing purchase intentions.

After conducting this research, it is necessary to stress that the effectiveness of message framing is most definitely context dependent and it is not a tool that can be implemented without analyzing the target group.

5.3 Limitations and further research

The size of the sample in this particular research was 165, for an even more representative research, and possibly more significant results a larger sample size would be advisable.

framing condition, which might influence the results. When taking the education levels into consideration, a large portion of the respondents participating in this research lean towards a high education level. A larger sample size might have helped in achieving a more equal distribution. Future research could look into the differences in regulatory focus between men and women.

Another factor limiting the current research is the fact that respondents might be biased regarding the valence of the message framing. Since respondents are exposed to one of the framing conditions and subsequently asked whether they find the valence of the product offering negative versus positive they might have a more detailed look at the product offering than they normally would. This implicates that they might have experienced the message framing even more positive or negative before answering questions regarding their purchase intention.

The dependent variable is measured using the items according to Bart et al. (2005). Even though the items seem solid for this research. There is some room for discussion whether the items only measure purchase intention. The second item may measure loyalty and the final two items of the purchase intention measurement might also measure attitude.

References

Andrade,& E.B.,& (2005).& “Behavioral+ consequences+of+Affect:+Combining+Evaluative+

and+Regulatory+Mechanisms”,&Journal(of(Consumer(Research,&32,&355Z362

Bart,&Y.,&Shankar&,&Sultan,& V.,&and&Urban,&G.L.,&(2005).& “Are+ the+drivers+and+role+for+

online+ trust+ the+ same+ for+ all+ websites+and+ consumers?+A+ large+ scale+ exploratory+

empirical+study.”&Journal(of(Marketing,&69:&133Z155

Block,&L.G.,&and&Keller,&P.A.,&(1995).&“When+to+Accentuate+the+Negative:+The+Effects+

of+ Perceived+ EfHicacy+ and+ Message+ Framing+ on+ Intentions+ to+ Perform+ a+ HealthL

Related+Behavior,”&Journal(of(Marketing(Research,&32,&192Z203

Cacioppo,& J.T.,& Petty,& R.E.,& (1982),& “The+ Need+ for+ Cognition,”& Journal( of(

Personality(and(Social(Psychology,&42:&116Z131

Cacioppo,&J.T.,&Petty,&R.E,.&Schumann,&D.,&(1983),+“Central+and+Peripheral+Routes+to+

Advertising+ Effectiveness:+ The+ Moderating+ Role+ of+ Involvement.”& Journal( of(

Consumer(Research,&10,&135Z

Cortina,&Jose&M.&(1993).&“Interaction,+nonlinearity+and+multicollinearity:+

Implications+for+multiple+regression,”&Journal(of(Management.

Chaiken,&S.&and&Maheswaran,&D.,&(1994),&“Heuristic+Processing+can+Bias+Systematic+

Processing:+ Effects+ of+ Source+ Credibility,+ Argument+ Ambiguity,+ and+ Task+

Importance+ on+ Attitude+ Judgment,”& & Journal( of( Personality( and( Social(

Psychology,&66:&460Z473

Chattopadhyay,& A.& and&Alba,& J.W.,& (1988),& “The+ Situational+Importance+ of+Recall+

and+ Inference+ in+ Consumer+Decision+ Making,”& &Journal(of( Consumer(Research,&

15:&1Z11

Cohen,& J.B.& and& Andrade,& E.B.,& (2004),& “Affective+ Intuition+ and+ TaskLContingent+

Affect+Regulation”,&Journal(of(Consumer(Research,&358Z368

Fennis,& B.& And& Stroebe,& W.& (2010),& “The+ Psychology+of+Advertising.”& Hove,& East&

Essex:&Psychology&Press&

Forsythe,&S.M.& and&Shi,&B.,& (2003),& “Consumer+Patronage+ and+ Risk+Perceptions+in+

Internet+Shopping,”&Journal(of(Business(Research,&56:&867Z875

Ganzach,& Y.& and& Karsahi,& N.& (1995).& “Message+ framing+ and+ Buying+ behavior:+ A+

Grant,& R.& Clarke,& R.J.,& Kyriazis,& E.,& (2007),& “& A+ review+ of+factors+ affecting+ online+

consumer+ search+ behavior+ from+ an+ information+ value+ perspective,”& Journal( of(

Marketing(Management,&23:&519Z533

Greenwald,& A.G.,& and& Leavitt,& C.,& (1984),& “Audience+ Involvement+ in+ Advertising:+

Four+levels,”&Journal(of(Consumer(Research,&11:&581Z592

Grewal,&D.,&Gotlieb,&J.,&Marmorstein,&H.&(1996),+“The+moderating+effects+of+message+

framing+ and+ source+ credibility+on+the+ priceLperceived+risk& Relationship,”!Journal(

of(Consumer(Research,&21:&145Z153.

Greifender,& R.,& Bless,& H.,& and& Pham,& M.T.,& (2011),& & “When+ do+ People+ Rely+ on+

Affective+and+ Cognitive+Feelings+in+Judgment?+A+ Review,”&Personality( and(Social(

Psychology(Review,&15(2):&107Z141

Hawkins,& S.A.,& Hoch,& S.J.,& and& MeyersZLevy,& J.& ,& (2001),& “LowLinvolvement+

Learning”,&Journal(of((Consumer(Psychology,&11:&1Z11

Herr,& P.M.,& Kardes,& F.R.,& and& Kim,& J.,& (1991),& “Effects+ of+ WordLofLMouth+ and+

ProductLAttribute+ Information+ on+ Persuasion:+ an+ AccessabilityLDiagnosticity+

Perspective,”&Journal(of(Consumer(Research,&17:&454Z462

Higgins,&E.T.,&Friedman,&R.S.,&Harlow,&R.E.,&Chen&Idson,&L.,&Ayduk,&O.L.,&and&Taylor,&

A.,& (2001),& “Achievement& Orientations& from& Subjective& Histories& of& Succes:&

Promotion& Pride& Versus& Prevention& Pride,”& European& Journal& of& Social&

Psychology,&31:&3Z23

Kahre,& A.,& Labrecque,& L.I.,& and& Asare,& A.K.,& (2011),& “The& Assimilative& and&

Contrastive&Effects&of&WordZOfZMouth&Volume:&An&Experimental&Examination&of&

Online&Consumer&Ratings,”&Journal(of(Retailing,&87:&111Z126

Keller, P.A., Lipkus, I.M., and Rimer, B.K. (2003). “ Affect, Framing, and Persuasion,” Journal of Marketing Research, 54-64

Klein,& L.,& (1998),& “Evaluating& the& potential& of& interactive& media& through&a& new&

lens:& search& versus& experience& goods,”& Journal( of( Business( Research,& 41:&

195Z203

Loken,& B.,& (2006),& “Consumer+ Psychology:+ Categorization,+ Inferences,+ Affect,+ and+

Lee,& A.Y.& and&Aaker,& J.L.,& (2004),&“Bringin+ the+ Frame+ Into+ Focus:+The+ InHluence+ of+

Regulatory+ Fit+ on+ Processing+ Fluency+ and+ Persuasion,”+ Journal( of( Personality(

and(Social(Psychology,&86:&205Z218

Maheswaran,& D.& and&MeyersZLevy,& J.,&(1990),&“The+InHluence+ of+Message+Framing+

and+Issue+Involvement,”(Journal(of(Marketing(Research,&27:&361Z367

Malhotra,& N.& K.& (2007),& “Marketing& Research,& An&Applied&Orientation”,& Pearson&

Prentice&Hall,&sixth&edition.

MeyersZLevy,& J.& and& Maheswaran,& D.& (2004).& “Exploring+ Message+ Framing+

Outcomes+When+Systematic,+Heuristic,+or+Both+Types+of+Processing+Occur,”+Journal(

of(Consumer(Psychology,&14&(1&2):&159Z167

Menard,& S& (1995).& Applied& logistic& regression& analysis.& Sage& University& paper&

series&on&quantitative&applications&in&social&sciences,&07Z106.&Thousand&Oaks,&CA:&

Sage.

Montgomery,&A.L.,&Li,&S.,&Srinivasan,&K.,&and&Liechty,&J.C.,&(2002),&“Predicting+Online+

Purchase+ Conversion+ using+ Web+ Path+ Analysis,”& Marketing( Science,& November&

2002

Mourali,& M.,& Böckenholt,& U.,& Laroche,& M.,& (2004).& “Compromise+ and+ Attraction+

Effects+ under+ Prevention+ and+ Promotion+ Motivations,”& Journal( of( Consumer(

Research,&34:&234Z247

Myers,& R& (1990).& Classical& and& modern& regression& with&applications& (2nd&ed.).&

Boston,&MA:&Duxbury

Vakratsas,&D.&and&Ambler,&T.,&(1999),&“How+Advertising+Works:+What+Do+We+Really+

Know?”,+Journal(of(Marketing,&63:&26Z43

Pham,& T.M.& and&Avnet,& T.,& (2004),& “Ideals+and+ Ought’s+and+the+ Reliance+ on+ Affect+

versus+Substance+in+Persuasion,”&Journal(of(Consumer(Research,&30:&503Z518

Rothman,& A.J.,& Bartels&R.D.,& and& Walschin,& J.,& (2006),& “The+ Strategic+ Use+ of+GainL+

and+LossLFramed+Messages+to+ Promote+Healthy+Behavior:+How+theory+can+Inform+

Practice?,”+Journal(of(Communication,&56:&202Z220

ZZZZZZZZZZZZZZZZ& and& Salovey,& P.,& (1997),& “Shaping+ Perceptions+ to+ Motivate+ Healthy+

Shen,&H.,&Wyer,&R.S.,&(2008),&“The+Impact+of+Negative+Affect+on+Responses+to+AffectL

Regulatory+Experiences,”&Journal(of(Consumer(Psychology,&18:&39Z48

Shiv,& B.,& Edell& Britton,& J.A.,& Payne,& J.W.,& (2004),& “Does+ Elaboration+ Increase+ or+

Decrease+ the+ Effectiveness+ of+ Negatively+ versus+ Positively+ Framed+ Messages?,”&

Journal(of(Consumer(Research,&31:&199Z208

Smith,& G.E.& (1996),& “Framing+ in+ Advertising+ and+ the+ Moderating+ Impact+ of+

Consumer+Education,”+Journal(of(Avertising(Research,&36&(5):&49Z64

Smith,&S.M.&and&Petty,& R.E.,&(1996),& “Message+ Framing+and+ Persuasion:+A+Message+

Framing+Analysis,”(Personality(and(Social(Psychology(Bulletin,&22:&257Z268

Wheatly,& J.J.& and& Oshikawa,& S.,& (1970),& “The+ Relationship+ between+ Anxiety+ and+

Positive+ and+ Negative+ Advertising+ Appeals,”+ Journal(of( Marketing( Research,& 7:&

85Z89

Yoo,& C.Y.,& (2005),& “Interplay+of+message+ framing,+keyword+ insertion+ and+ levels+of+

product+ involvement+ in+ clickLthrough+ of+ keyword+ search+ ads,”& International(

Journal(of(Advertising,&30:&399Z424

Zhang,& Y.& and& Buda,& R.,& (1999),& “Moderating+ effects+ of+ Need+ for+ Cognition+ on+

Responses+to+Positively+versus+Negatively+Framed+Advertising+Messages,”+ +Journal(

Appendices

III Reliability tests

Construct Items (7-point likert scales)

_∝

deal if you make a mistake

0.613 Bearden, Haws and Netemeyer, 2001

2. It is really annoying to purchase a house that is not suitable

3. If, after I bought a product of Kadaster, my choice(s) prove to be poor, I would be really upset.

4. Whenever one buys a house, one never really knows whether it is the one that should have been bought.

5. When I face prospect houses, I always feel a bit lost.

6. Choosing a house is rather complicated. 7. When one purchases a house, one is never certain of their choice.

8. You can tell a lot about a person by the house he or she chooses

9. The house I buy gives a glimpse of who I am as a man/woman

10. The house you buy tells a little bit about you

11. It gives me pleasure to buy a house 12. Buying a house is like buying a gift for myself

13. Buying a house is somewhat a pleasure to me

14. I attach great importance to choosing an information product from Kadaster 15. One can say real estate interests me a lot

16. Real estate is a topic which leaves me totally indifferent

Promotion focus

1. Compared to most people, are you typically unable to get what you want out of life?

0.610 Higgins et al. 2001

7. Do you often do well at different thing that you try?

3. How often have you accomplished things that got you “psyched” to even work

harder?

9. When it comes to achieving things that are important to me, I find that I don’t perform as well as I ideally would like to do 10. I feel like I have made progress toward being successful in my life

11. I have found very few hobbies or activities in my life that capture my interest or motivate me to put effort into them Prevention

focus

2. Growing up, would you ever cross the line by doing things that your parents would not tolerate?

0.811

8. Not being careful enough has gotten me into trouble at times

4. Did you get on your parents nerves often when you were growing up? 5. How often did you obey rules and regulations that were established by your parents?

6. Growing up, did you ever act in ways that your parents thought were

objectionable? Purchase

intention

1. I would purchase an item at this site 0.677 Bart et al. 2005 2. I would recommend this site to a friend

3. I am comfortable providing financial and personal information on this site

4. I would book mark this site Manipulation

Check

I think the valence of this product offering is (positive/negative)

The product offering on this web page emphasizes on the negative consequences avoided when one buys the product

The product emphasizes on the losses incurred when one does not buy the product

IV SPSS Output - Manipulation Check

VI!

!

Independent Samples t-Test - Age