The procedure by which a few subjects are chosen from the *universe* to be studied in such as way that the *sample* can be used to estimate the same characteristics in the total is referred to as sampling. The advantages of using samples rather than surveying the *population* are that it is much less costly, quicker and, if selected properly, gives results with known accuracy that can be calculated mathematically. Even for relatively small samples, accuracy does not suffer even though precision or the amount of detailed information obtained, might. These are important considerations, since most research projects have both budget and time constraints.

Determining the population targeted is the first step in selecting the sample. This may sound obvious and easy, but is not necessarily so. For instance, in surveying "hotel guests" about their experiences during their last stay, we may want to limit the term to adults aged 18 or older. Due to language constraints, we may want to include only those guests that can speak and/or read English. This more operational definition of population would be considered the "respondent qualifications". They make the research more doable, but also introduce delimitations of the study’s scope.

Next, the *sampling units* themselves must be determined. Are we surveying all hotel guests that fit our operational definition, one person per room occupied, one person per party (and who?), households, companies, etc.?

The list from which the respondents are drawn is referred to as the *sampling frame or working population*. It includes lists that are available or that are constructed from different sources specifically for the study. Directories, membership or customer lists, even invoices or credit card receipts can serve as a sampling frame. However, comprehensiveness, accuracy, currency, and duplication are all factors that must be considered when determining whether there are any potential *sampling frame errors*. For instance, if reservations and payments for certain business travellers is made by their companies without specifying the actual guest name, these would not be included if the sampling frame is the hotel’s guest list. This could lead to potential underrepresentation of business travellers.

Please see these two helpful articles by Pamela Narins from the SPSS site about calculating the Survey Sample Size and correcting the Finite Population. Also, this article by Susanne Hiller from the National Post highlights the problems you can run into in interpreting survey results when your sample size is too small. In this article published in the Globe and Mail on Consistent Mistakes that Plague Customer Research, George Stalk, Jr. and Jill Black discuss common problems and their implications. Reuter's Washington office reported on the misuse of polls and how results can be biased because of the question wording in this Toronto Star article dealing with the Microsoft case.

In *probability sampling*, the sample is selected in such a way that each unit within the population or universe has a known chance of being selected. It is this concept of "known chance" that allows for the statistical projection of characteristics based on the *sample* to the *population*.

Most estimates tend to cluster around the true population or universe mean. When plotted on a graph, these means form what is called the *normal or bell curve.* This theoretical distribution allows for the calculation of the probability of a certain event occurring (e.g. the likelihood that an activity studied will be undertaken by people over 65 years old, if those are the variables being studied).

There are three main types of probability or random sampling that we will review more closely:

In *non-probability sampling*, the sample is selected in such a way that the chance of being selected of each unit within the *population or universe* is unknown. Indeed, the selection of the subjects is arbitrary or subjective, since the researcher relies on his/her experience and judgement. As a result, there are no statistical techniques that allow for the measurement of *sampling error*, and therefore it is not appropriate to project the sample characteristics to the population.

In spite of this significant shortcoming, non-probability sampling is very popular in hospitality and tourism research for *quantitative research*. Almost all qualitative research methods rely on non-probability sampling techniques.

There are three main types of non-probability sampling that we will review more closely: