Research Methods in Tourism, Hospitality and Events Management

4.4.2 Random sampling

Random sampling is almost the same as probability sampling with one slight

exception, which is pertinent to the topic of this book. As has been shown, there may

be situations where the population in a survey is not known. With the case of

informal recreation sites or events, which are admission-free, this is often the case,

with the total number of visitors from year to year being estimated rather than

calculated. Hence, true probability sampling is not possible. The principle of random

selection can still apply, but the probability of inclusion in a survey cannot, as the

population size is unknown. Take, for example, interviewing, on a next-to-pass basis,

200 people leaving a beach. They are selected on a random basis but it is not known

whether the 200 interviewed are from a population of 1,000, 1,500 or 5,000 beach-

goers on that day or over a season. Awareness of the limitations this may cause in

terms of the inability to predict from these 200 the attitudes of all visitors to the

beach is clear. However, this should not suggest that such a survey would be

worthless, as descriptive studies or ‘straw polls’ can be of much value in certain

situations where predictions and generalisations about the rest of the population are

not necessary.

Nevertheless, probability sampling, where a population is known, can be seen as

preferable wherever it is possible. This is because statistical techniques can be used

to suggest the amount of ‘error’ between the sample and its representativeness of the

rest of the population. This specific aspect will be investigated in a later section.

Returning to the different types of sampling, it can be seen that probability sampling

and random sampling share many characteristics. Both methods involve the random

selection of sampling units in a survey. Because they are so similar in many respects,

the specific difference is often ignored and either the term ‘probability sampling’ is

loosely used to include ‘random sampling’ or vice versa. Having defined the starting

point, it is now possible to look in more detail at some more specific types of

sampling methods which are based on the principles of equal chance and random

selection.