• No se han encontrado resultados

1º.- MARCOS ENZO TALLARITA – 28459/1 -

In document CONCEJAL Alcolea Hernan (página 22-25)

The coding schemes were developed from four sources: (i) the research questions; (ii) passenger progress through the airport; (iii) actual observation of how

Chapter 5 Research design 57

passenger activities enable their airport experience; and (iv) the literature on the airport experience. While the coding schemes used in Field Studies One and Two have the same foundation, each study has a different focus. The full coding schemes for each field study are explained and discussed in their respective chapters (Chapter 6 and 7 Field Study One; and Chapter 8 Field Study Two).

There are three levels of coding used for all data in Field Study One. They are (i) macro level coding, (ii) location coding, and (iii) activity coding. Table 5.1 illustrates these levels. The activity coding level only was used for Field Study Two.

Table 5.1: Coding scheme levels for the two field studies

Coding name Number of codes

Used to code

Macro 2 Processing and discretionary divide Location 18 The location of the passenger

Activity 30 The activity undertaken by the passenger

Macro level coding

All data was coded for what has been termed the “macro level” of the airport experience. This level is simply divided into “processing” and “discretionary” (Kraal et al., 2009; Popovic et al., 2009). These two levels of coding are introduced and explained in Section 2.4, and illustrated in Figure 2.4. Activities were coded as

“processing” when passengers entered one of the domains associated with processing: check-in, security, or customs. Processing at these domains was location dependant. Processing was coded for boarding when the flight was called by the airline. Therefore, coding processing at boarding was not location dependant. Coding of processing in the interview transcripts occurred when a passenger or staff member discussed one of the processing domains: check-in, security, customs, or boarding.

Discretionary periods were coded in Noldus Observer once passengers had finished their processing activities. Once they left the check-in desk or walked through the security and customs area, they were coded as being in a “discretionary”

period. Coding of discretionary periods in the interview transcripts occurred when a passenger discussed an activity that was not associated with processing.

58 Chapter 5 Research design

Location level coding

The next level of coding was “location” and all raw data was coded at this level. The coding of locations was dependent on the airport being observed. The three processing domains were coded when a passenger entered a specific location.

All three domains had physical borders at every airport, such as bollards, doors, floor markings, or a change in level. These physical borders allowed clear boundaries when coding locations. This was consistent at the three airport locations.

All the airports used in the project had at least five different types of shops;

these included duty free outlets and newsagencies. In an attempt to achieve consistency, types of shops were not coded – “Shops” was the term used for all retail outlets that sold predominantly non-food. All food outlets were coded as “cafe” – the term used for all retail outlets that sold predominantly food – and all airports had at least five different food outlets. Coding of “cafe” included areas in close proximity to the cafe where passengers could sit and eat.

Coding of “seating area” referred to areas where passengers could sit but which were not associated with any retail outlets or located near a departure gate. If the seating area was near a departure gate, it was coded as “departure gate”. It is important to note that passengers could be at the departure gate, but not boarding. As discussed above “boarding” only occurs when the airline calls the passenger to board their flight; this is why boarding is not coded as a location. The final code was

“amenities”. Coding for this started when the passenger approached the corridor leading to a toilet, and continued until they were seen leaving the same corridor.

Locations were coded in the interviews when they were discussed by the passenger or staff member.

Activity level coding

The most detailed level coding of observational data was activity. Activity codes were determined by watching the video footage and coding what activities the passengers actually undertook during observation. Coding was mostly done through watching the videorecording. However, on occasion, the video camera was obscured by another person, or was prohibited; for example, at customs. Activity coding was determined by the researcher’s spoken description that was recorded during the videorecording. Coding arising from the interviews was determined by the description given in the retrospective interview transcripts.

Chapter 5 Research design 59

Research rigour

To ensure there was no researcher bias, 20% of the videos and 20% of the interviews were re-coded by a ‘blind’ researcher who had limited knowledge of the respective coding levels used. When the fourteen videos were re-coded, there was complete correspondence between the original coding and the blind re-coding for the

“macro” and “location” levels. There was a slight difference in the timing of when the activities started and stopped; however, this was only a difference of seconds.

The total number of times that activities were undertaken was the same. The difference between the coding and re-coding was less than 5%.

The interviews were also coded by a ‘blind’ researcher. There were only small differences between the blind researcher’s coding and that of the original researcher.

These differences were resolved by making the definitions of the coding levels clearer. Again, the difference between the two coders was less than 5%.

In document CONCEJAL Alcolea Hernan (página 22-25)

Documento similar