**3. Methodology**

Using a qualitative study, we sought to understand how consumers categorize a new hybrid electronic device on the basis of its affordances and how the latters affect the level of perceived risk. The product under study was a new concept of smartphones designed by Kristian Larsen Ulrich, called the "Flip Phone." This original concept of smartphone has three touch screens with a foldable keyboard and a camera (see **Figure 1**). This concept was chosen as it illustrates a hybrid product combining diverse base categories (telephone, camera, tablet, and netbook, etc.) with particular form and design. The fact that it is not yet commercialized ensures that there is no prior categorization in the mind of our research sample.

#### **3.1 Data collection**

Twenty-four semi-structured and thematically focused interviews were conducted until theoretical saturation was reached. Each interview lasting approximately

**Figure 1.** *The Flip phone concept.*

45 minutes was recorded and transcribed. The sample aimed to cover the largest variety of consumer profiles. As we were studying evaluation and categorization processes based on a high-technology product, the selection criteria were based on age, gender, and revenue, as well as the possession (or not) of some specific electronic devices (e.g., smartphone and tablets). The final sample consisted of thirteen men and eleven women with an average age of thirty-nine years ranging from eighteen to seventy-eight years old (see the respondents' profiles in **Table 1**) and a wide range of occupations to cover the different social positions that can promote or constrain the acquisition of this type of electronic device.

The study was conducted in two phases: first, photos of the Flip Phone, without any additional written information, extracted from the designer's website were shown to respondents who were asked then to guess what this item might represent, identify its potential uses and try to evaluate its performance derived from its sole ergonomy and design. In a second step, subjects were asked to attribute it freely to a product category either existing one or imagined/to be created one and to explain in few


**Table 1.** *Informant information.* *Design Affordance Does Matter: A Spotlight on Categorization and Evaluation of Hybrid… DOI: http://dx.doi.org/10.5772/intechopen.112966*

sentences the process they adopted to associate it with that particular category. The categorization was inferred using a behavioral indicator suggested in [5], namely, the intention of the respondent to "look for the NHP device at a store or a website within a department corresponding to one or the other of its basic categories, or in another department — to be defined (if affected to none of them)."

#### **3.2 Reliability and validity of data analysis**

We applied several criteria of research trustworthiness (such as integrity, confirmability, and transferability) as a means of ensuring the quality of the research methodology [26]. The integrity of interpretations was taken into account avoiding any misinformation or evasions by participants. Interviews were conducted in a nonthreatening way. No participant stopped the interviews or expressed any concern about the line of questioning. The confirmability of the interpretation was assessed as the coding process was conducted by two different researchers. Intra- and inter-coder reliability were undertaken: (1) After the initial coding of the data, one of the authors repeated a month later the coding of the same discourse material attaining 95% of intra-coder reliability. (2) Another subset of interviews (20% of the whole material) was double-coded by two coders trained in qualitative data analysis. This subset size can be considered, according to [27], as highly sufficient in a double coding procedure, especially when the part of interviews under scrutiny is quite varied and rich. Differences between coders were discussed until agreement was reached as outlined by [28]. We achieved an inter-coders agreement rate of 83% by counting the number of agreements and dividing it by the sum of agreements and disagreements, which attests to the sufficient reliability of our analysis.

Coding was iterative; during the process, we complemented the initial inductive codes with those derived from literature then we refined them as we collected more data [28]. The coding of the transcripts and thematic content analysis was conducted using manual analysis and then NVivo software in a validity perspective. Furthermore, the transferability of the findings may be ensured to other hybrid products combining the same electronic base categories.
