Research Methods in Tourism, Hospitality and Events Management

Example: travel agent services

Background

A branch of a travel agency has been established for the first time in a town. A

survey is decided to ascertain the awareness of this venture amongst the resident

population of the surrounding area, but no complete list of all residents is available.

Stages

1. A map is used to define the boundary of potential users of the shop.

2. The area within this boundary is divided into segments, reflecting all areas, both

residential and non-residential.

---

3. Each segment is numbered.

4. A sample of segments is selected on a simple random or systematic basis.

5. Within the segments, actual residences are numbered.

6. Systematic or simple random samples of residences are selected.

7. Individuals within the selected residences are contacted for interview.

In this example, two clusters are used: segments within the defined boundary and

residences within each segment. In some surveys, it may be necessary for several

stages of clustering to take place in this way. Where this occurs, the process is

sometimes referred to as ‘multi-stage sampling’.

A variation on this theme can be found with what is termed ‘multi-phase sampling’;

the difference between ‘multi-stage’ and ‘multi-phase’ often relates to the latter

using a follow-up survey. In our example, a first ‘phase’ might be to investigate the

awareness of the new travel agency outlet. From the results of this, a second ‘phase’

could be to develop a follow-up survey amongst those respondents who answered a

particular question in a specific way. For example, further details could be sought

from those who have indicated that they are planning to book a holiday in the next

six months, or those who have asked to be informed of special offers. Thus, multi-

phase sampling could be used to first gain a general picture and then to follow up

with more detailed (and longer) interviews in a subsequent phase with particular sub-

sets of respondents. In all phases, the respondents are always selected randomly.