The aim of this study is to analyse consumers’ attitudes towards two hypothetical products licensed by Pringles. We have employed quantitative approach and in this section we will present and discuss how the data collection design was determined. Questionnaire implementation, analyses techniques and quality of the chosen method will also be presented to the reader.
Whilst conducting empirical research there are two possible research methods – quantitative and qualitative (Saunders, Lewis, & Thornhill, 2009). The underlying difference between the two types is that the quantitative approach focuses on obtaining a large sample that is then further analysed using statistical methods. On the contrary, qualitative approach enables the researcher to get a better understanding of the situation or a problem and is more descriptive in nature (Malhotra, 2004).
To illustrate this, one can use McGrath, Martin, and Kukla (1982) parable, which states that quantitative research is a picture with the bird‟s eye of view, meanwhile the qualitative research shows all the details of the picture. A third way to differentiate between the two techniques is to refer to the quantitative approach as a synonym for data analysis that produces numerical data. Alternatively, qualitative techniques produce non-numerical data (Saunders et al., 2009).
Whether the quantitative or the qualitative method is the superior one has been previously discussed in the literature (Malhotra, 2004; Zikmund & Babin, 2010). According to Zikmund and Babin (2010), both methods have their advantages, but what is the most important is to match the right approach to the right research context. The qualitative approach is appropriate when the aim is to understand and describe consumers underlying reasons and motivations, meanwhile the quantitative approach seeks to quantify data (Malhotra, 2004). Furthermore, the qualitative approach is considered more subjective since the researchers are highly involved in the process (Zikmund & Babin, 2010). The quantitative approach is found to be more objective since the respondents provide the answers and the researchers are uninvolved (Zikmund & Babin, 2010).
The purpose of this study is to investigate consumer‟s attitude towards licensed products in relation to the parent brand, with respect to perceived quality, likelihood to buy and associations‟ transferability. Accordingly, the quantitative technique is appropriate to answer the purpose of this study as the aim is to describe existing attitudes and do not go into detail why these attitudes exist. The numerical nature of the technique allows for hypotheses testing and enables clear illustration of relationships in the form of charts, diagrams and statistics (Zikmund & Babin, 2010). The mentioned tools allow to discover relationships and trends within the data collected (Saunders et al., 2009).
The research “onion”
During the design of our research, we have used the “onion” approach developed by Saunders et al. (2009). The research onion categorises possible approaches, strategies and techniques available to conduct research. According to the approach, the researchers start with selecting a broad research approach and then further “peel” the other layers to design the data collection.
We will present the research approach and the research strategy used to fulfil the purpose followed by the time horizon as well as the data collection method. Furthermore, we will describe our choice of sample, approaches used during the analyses of empirical data and, finally, study‟s method quality will be discussed.
When conducting research, there are two fundamental approaches that can be used to gather data – deductive and inductive approaches (Burns & Burns, 2008; Saunders et al., 2009).The main difference between the two is that the inductive approach works through building a theory and the deductive approach focuses on testing a theory or hypotheses (Saunders et al., 2009). When the inductive approach is used, the theory follows data unlike with the deductive approach (Burns & Burns, 2008). Our aim was to test hypotheses, hence, deductive approach was more relevant to fulfil the purpose of our study. In our study we tested if certain relationships between two licensed products and the parent brand would hold. As such, the deductive approach aims to find underlying relationships between variables. When employing the deductive approach, the researcher stays independent and objectively collects data using structured methodology (Saunders et al., 2009). This allows other researchers to replicate the study and increases study‟s generalizability.
“Surveys gather information to assess consumer knowledge and awareness of products, brands, or issues and to measure consumer attitudes and feelings”
(Zikmund & Babin, 2010, p.146)
We have used questionnaires to collect the results for our study. When gathering information, one can also use field studies, observations, interviews and focus groups (Saunders et al., 2009). Questionnaires helped us to answer the stated research questions as questionnaires are used to identify what people believe, attitudes towards variables and finding relationships between variables (Zikmund & Babin, 2010).
According to Zikmund and Babin (2010), there are two approaches to administering questionnaires – interactive or non-interactive approaches. The main difference between the two is that interactive approach allows for instant two way interaction. Self-administered questionnaires were used to collect the data. In a traditional sense, self-administered questionnaires use non-interactive approach, however, the researchers
were present when respondents answered the questionnaires and the respondents were explicitly told that they could ask for any clarifications they need.
When participants answer using self-administered questionnaires they are themselves responsible for reading and answering the questions (Zikmund & Babin, 2010). The main reason for using self-administered questionnaires is that it allowed us to give out several questionnaires at the same time, thus saving time. Alternatively, we could have administered the questionnaires through a personal interview, by asking the questions and then writing the answers. This does have the advantage of face to face exchange resulting in more effective communication (Zikmund & Babin, 2010). If the interviewer feels that the respondent does not fully understand the question, clarification can be provided. Moreover, clarification from the respondent can be asked if the answer is too vague, this is known as probing (Easterby-Smith, Thorpe, Jackson, & Lowe, 2008). However, there are some limitations to this way of administering questionnaires. Interviewer bias is one of the biggest disadvantages of personal interviews (Zikmund & Babin, 2010). Different interviewer techniques, age and even gender can have adverse effects on the answers (Zikmund & Babin, 2010). Conducting questionnaires using interviews are also more time consuming and do not allow to conduct it with more than one participant at a time. For the mentioned above reasons this method was deemed as not feasible for our study.
In general, questionnaires are a practical way of collecting data, as it is quick, easy and not as expensive as other methods of data collection (Zikmund & Babin, 2010). The data is standardised which allows for comparison between cases. Further, statistics can be used to analyse the data. However, no single method is perfectly suited for conducting any particular research. When collecting data with questionnaires, one of the biggest drawbacks is that respondents may not be willing to provide information, resulting in a low response rate (Malhotra, 2004).
Rejected data collection methods
Before moving on to the specific questionnaire design, we would like to specify why other data collection methods were decided not to be feasible options.
We have considered using focus groups to collect data needed to answer the purpose. However, in addition to this method being qualitative, there were several other reasons why this method was rejected. The main limitation of using focus groups is that it requires objective and well-trained moderators (Zikmund & Babin, 2010). If more than one moderator is used, then they have to perform the task equally across different groups. Because we were not trained in conducting focus groups, there was a high chance that it would not produce unbiased and valid responses. Moreover, focus groups are very difficult to moderate (Malhotra, 2004). Some focus group participants may dominate the session by giving their opinions and suppressing others (Zikmund & Babin, 2010). Finally, the costs were also considered, as the respondents would have to be compensated for their time.
Interviews were also considered as a way of collecting data to answer the study‟s purpose. In depth interviews are considered an alternative to focus groups (Zikmund & Babin, 2010). Once again, the interviewer‟s skill and training is vital to properly conducting interviews. The interviewer must not affect the respondent‟s answers. Interviews have very similar limitations and disadvantages to focus groups (Malhotra, 2004). In addition, the data gathered is difficult and very time consuming to analyse.
Questions can be designed to be either structured or unstructured (Malhotra, 2004). Our questionnaire is designed using both types and open ended questions, noncomparative scales – namely, the Likert scale and list type questions. The questionnaire was designed to only collect data that is required to answer the purpose. Open-ended questions allowed the respondents to list the associations towards Pringles and two hypothetical, licensed products using their own. This is an example of an unstructured question (Malhotra, 2004). Such type of questions produces more valid results, because the respondents are not forced to choose one of our predetermined answers (Malhotra, 2004). Furthermore, when using open-ended questions, cultural bias is reduced. This is because researchers may have different cultural opinions and provide selectable answers that are not representative of respondents‟ true feelings (Kumar, 2000). List question is an example of a structured question, it offers respondents a pre-determined answer (Saunders et al., 2009). This type of questions was used to answer respondents‟ age, gender and country of origin.
Noncomparative scale questions were used to measure respondents‟ attitude towards the three products by evaluating one subject at a time (Malhotra, 2004). This was done using the Likert scale, when the respondents were provided with a statement, followed by five response options, indicating their level of approval of the sentence (DeVellis, 2003). DeVellis (2003) states that Likert scale is commonly used to measure attitudes and beliefs. One of the biggest advantages of using Likert scale is that the answers are not based on subjective opinions and the Likert scale provides the respondents with a homogenous scale (Burns & Burns, 2008). The respondents find it easy to understand Likert scale questions and it is one of the most common scales used in marketing research (Zikmund & Babin, 2010). According to DeVellis (2003), Likert scale can contain either odd or even number of responses. We have used a five point Likert scale; this is one of the most common ranges used by researchers (Burns & Burns, 2008; Zikmund & Babin, 2010).
Moreover, having fewer possible replies to a question, such as five, reduces the complexity of a question, which is particularly important in self-administered questionnaires, as they have to be easy and encouraging to answer to increase the response rate (Zikmund & Babin, 2010). The use of this range also allowed the respondents to stay somewhat neutral by choosing the middle option (DeVellis, 2003). When expressing a negative attitude respondents could choose either option one or two and provide a positive reaction by choosing four or five. We have also followed guidelines by DeVellis (2003) who recommends wording the response options with equal distance between them, so this equality is apparent to respondents.
A total of 13 questions were asked, Table 3-1 represents these questions, what information was gathered from each question and what theory was used to analyse it. Wording the questions is one of the most critical parts of designing a questionnaire (Malhotra, 2004). The wording of the questions was carefully chosen to produce simple, not leading and unbiased questions. It was particular important to have the respondents understand the question in the same way as we do, to produce valid and reliable results (Zikmund & Babin, 2010). We have used techniques presented by Oppenheim (1992) to achieve clear and easy to understand questions: long, complex questions were avoided, no questions with double meaning were included and no leading questions were asked. According to these techniques, emotion and double meaning words should not be used in the wording of questions (Oppenheim, 1992).
The questionnaires have been administered in person and over the internet using “Google Docs”. All respondents were briefed prior to filling in the questionnaire by means of a cover letter (Appendix A). It informed the respondents that all of the answers will be treated confidentially and that the data was being collected for a bachelor thesis. Previous research has shown that a cover letter can persuade the respondent to answer the survey (Saunders et al., 2009; Zikmund & Babin, 2010). Therefore, it was deemed important to have an individual cover letter.
During the administration of questionnaires, we had followed ethical guidelines developed by Saunders et al. (2009). In particular, privacy of all participants was respected by making the questionnaire anonymous. All of the participants volunteered to answer with no compensation provided and were told that they can withdraw from answering at any time. Once the data from the questionnaires was transferred to SPSS, the forms were destroyed.
The questionnaire was clearly divided into four sections: demographics, Pringles chips, Pringles dip and Pringles backpack. The last three sections were almost identical, except to the referral to the product. This was done to ensure the consistency of data collection.
Because self-administered questionnaires were used to gather data, it was particularly important to make the questionnaire look appealing and not long. Shorter questionnaires increase the number of replies and the dropout rate (Saunders et al., 2009). Grammar was carefully checked prior to administration for professionalism. The questionnaire was distributed on one A4 size page and all thirteen questions fit on one side (Appendix A). To reduce costs, the questionnaire was designed and printed in black and white. In order to increase readability and make the answers section stand out, questionnaires had grey shading. Using the introductory text, the participants were told that they would have to tick the most appropriate answer in the boxes provided.
“Page-by-page” layout was used for the questionnaire administered over the internet using “Google Docs”, each section was presented on a separate page. This significantly increases the skip patterns and the dropout rate (Zikmund & Babin, 2010). Moreover, the participants were required to answer every question, before they could proceed to the next page as every question is vital to answer the purpose of the study. The questions included in the online version of the questionnaire were identical to those distributed on campus. In both versions, the questions were grouped in sections described above.
The questionnaire was originally written in English. However, as we conducted our research on a Swedish sample, not all individuals possess a very high proficiency in the English language. Therefore, it was important to provide a Swedish version of the questionnaire. The main reason for Swedish translation was to help the respondents understand the questions and avoid errors because of language misinterpretation (Saunders et al., 2009).
As a result, the questionnaire had to be translated and it was particular important to assure the questionnaire had the same meaning to all respondents so it could be answered as intended (Saunders et al., 2009). According to Usunier (1998), four factors require special attention when translating a questionnaire:
Lexicon meaning – The exact meaning of individual words may differ between languages.
Idiomatic – Group of words that has a meaning to the native speaker but does not make sense as individual words.
Experiential meaning – The meaning of words and sentences may make sense to the respondent in one context but may not have an equivalent meaning in another context.
Grammar and syntax – Grammar and order of words can be significantly different in different languages.
Usunier (1998) describes a further four specific techniques of translating a questionnaire:
Direct translation – One person undertakes the translation. Easy and inexpensive to implement but has a high possibility for errors.
Back-translation – The original questionnaire is translated which is then followed by a back-translation to the original language. A comparison between the two is made and any errors detected.
Parallel translation – Two or more people independently translate the questionnaire and then the versions are compared to create the final version.
Mixed techniques – A back–translation performed by two or more independent translators where the versions are compared and combined into the final version.
The direct translation method was excluded because of its high possibility of error. Moreover, mixed and back-translation require more than two native speakers in each language and therefore were excluded. Due to the limited resources, combined with our aim to minimize possible errors during the translation the parallel translation technique was chosen. Two independent native Swedish speakers translated the original English questionnaire into Swedish. The two versions were then compared and combined; this has resulted in the final Swedish version that was distributed in person as well as through “Google Docs” (Appendix A).
Table of Contents
1.2 Problem discussion
1.4 Research Questions
2 Theoretical Framework
2.1 Choice of Theory
2.2 What is a Brand
2.3 Brand Extensions
2.4 Brand Equity Model
2.5 Reason to Buy
2.6 Brand Fit
2.7 Attitude towards brands and products
2.8 Summary of theory
3.1 Quantitative Approach
3.2 The research “onion”
3.3 Research Approach
3.4 Research Strategy
3.5 Time Horizon
3.6 Data Collection Methods
3.7 Analysis Approach
3.8 Method Quality
4 Empirical Findings
4.1 Study Demographics
4.2 Level of Awareness
4.3 Perceived quality
4.4 Likelihood to buy
4.5 Brand associations
4.6 Attitudes towards products
5.1 Perceived Quality
5.2 Likelihood to Buy
5.3 Brand Associations
5.5 Modification of Aaker‟s Model
5.6 Explanation of the model
7.1 Managerial Implications
7.2 Critique of the study
7.3 Other Findings
7.4 Suggestions for further research
GET THE COMPLETE PROJECT
Brand Licensing Once you pop you can‟t stop: When brand licensing goes too f