Data collection methods

Data collection of this study consist of primary and secondary data as formerly mentioned. It is essential to set a theoretical concept that is being evaluated in the context of the mobile channel and the collected data, to utilise abductive research approach to theory creation. This section shortly describes secondary data collection methods and then moves on to the primary data collection.

4.2.1 Secondary data

Secondary data was used to produce a literature review based on two topics of mobile as a shopping device and CE has been presented in chapters 2 and 3. The secondary data used was collected using journals and books available from Jyväskylä University library and articles available in electronic databases such as Google Scholar and ResearchGate. Table 3 presents the most important keywords utilised in this research.

TABLE 4.1 Keywords used in secondary research

Mobile as a shopping device Customer experience

• Mobile shopping

• Online shopping

• Mobile commerce

• Mobile retailing

• In-app retail

• Experimental marketing

• Online customer experience

• Mobile customer experience

• Customer experience

• User experience (UX)

As the abductive research logic is applied in this study, it is essential that the secondary and the primary data support each other, assisting the researcher to answer the research questions. Regarding explorative research design, literature review is typically shorter and used to clarify the lack of research in the field (Stebbins, 2001, 42). However, in this study, literature review is providing suggestions of what factors may be relevant to the mobile shopping CE and it is therefore more extensive.

From the basis of the literature review, a conceptual model was created. It is presented in Figure 5 and there are five main themes directly following the dimensions of CE discussed in Chapter 3. There are also multiple subthemes based on theoretical concepts from the previous literature around the topic. The primary data collection follows these themes identified from the literature review and utilises them as the priori coding scheme.

FIGURE 4.2 Conceptual framework

4.2.2 Primary data

Qualitative in-depth interviews were chosen as the most suitable method for attaining detailed data to convey thoughts and feelings that might affect the way participants behave (Sutton & Austin, 2015).

Sampling

Regarding the sampling, purposive and snowball sampling are utilised, and the sampling is based on a specific purpose rather than random selection (Jupp, 2006, 244). The sample consists of smart phone users who reside in Finland and are between ages 18 and 30, from generations Y and Z. The overall goal is to attain even gender and age divide within the sample. Snowball sampling is used to create multidimensionality to the sample by asking participants if they know someone who would be suitable participant for this study. Furthermore, whilst the sample size is limited to 6-10 interviews, this research aims to follow the principle of thematic saturation, whereby the adequate sample size is achieved when the interviews no longer reveal new themes or information (Guest, Namey

& Chen, 2020; Saunders et al., 2019, 206).

Interview protocol

Semi-structured interview protocol was designed, consisting of closed- and open-ended questions with appropriate prompts (Adams, 2015, 493). The protocol follows the guidelines of Castillo-Montoya (2016), ensuring that interview protocol is designed with the research questions in mind. Due to abductive research logic, the questions are based on existing theory and the conceptual model (Figure 5) that was built on the theoretical concepts introduced in the literature review. Table 4 presents some examples of the questions and on what theoretical concepts they are linked to. Full interview protocol can be found in Appendix 1. Length of the interviews is approximately 60 minutes to avoid tiring the interviewer or the participant and to keep the quality of collected data good. The interviews are recorded and manually transcribed for data analysis.

TABLE 4.2 Examples of interview questions

Do you think that mobile shopping could be efficient and convenient in achieving your shopping goals?

(Zheng et al., 2019) Have you ever begun using an app because you have been

recommended it?

Intellectual Creative use

(Brakus et al., 2009)

Do you enjoy finding new ideas, products and promotions through mobile device?

Back translation

It is important to take into consideration that this thesis written in English whilst the interviews are conducted in Finnish. To ensure the correct translation, back translation technique, as described by Kolb (2008, 75-76), was utilised. Therefore, interview questions were first written in English, then translated to Finnish by the researcher, and translated back to English by a colleague. These translations were then compared, and the researcher also consulted a native English speaker to ensure that the questions presented in Appendix 1 were an accurate presentation of the questions asked in the interviews. Whilst interviews are semi-structured emphasis is put on researcher’s understanding of the research questions and how much they can be rephrased in the social situation such as informal interview.

Pilot testing

Lavrakas (2008, 584) argues that pilot tests can be used to better understand the suitability of the interview questions in terms of their comprehensibility and sensitivity. Pilot testing is more typically used in survey research, where there is a need to better understand if the measurements of the constructs are correct and that there are no errors in survey design. In case of this qualitative study, it was important to conduct pilot testing for the researcher to evaluate the interview protocol as well as to practice conducting an interview. The researcher was able to see how interview protocol worked in practice and how the responses of the participants could affect the course of the interview. Particularly, elaborating the interview questions and providing suitable examples was put under examination.

Prior to conducting the interviews, the interview protocol was tested on two participants. After the pilot tests were conducted no major errors were detected.

Prompts and examples were added to make it easier for participants to understand and answer the questions. Some questions were slightly simplified to better fit the interview situation. Data collected during these pilot interviews was considered and included in the data set of the study.

Ethical considerations

In the current circumstances of Covid-19 epidemic, interviews were hold via Zoom at a distance to ensure safe participation. Data privacy and research notifications (Appendix 2) were issued and discussed with the participant prior to the interview. Interviews were recorded and the recordings were deleted after completion of manual transcription. Each participant and transcription were assigned a number to enhance privacy whilst still enabling withdrawal from the study. Interview process and questions were not planned to be invasive or harmful in anyway but in case the participant feels uncomfortable in the interviewing situation, they have been encouraged to leave the Zoom.