• toggle

• toggle

• toggle

• toggle

# Handbook of Research Methods in Tourism

## Edited by Larry Dwyer, Alison Gill and Neelu Seetaram

This insightful book explores the most important established and emerging qualitative and quantitative research methods in tourism. The authors provide a detailed overview of the nature of the research method, its use in tourism, the advantages and limitations, and future directions for research.

# Chapter 1: Statistical Testing Techniques

## Gang Li

Subjects: development studies, tourism, environment, environmental sociology, tourism, geography, tourism, research methods, qualitative research methods, quantitative research methods

## Extract

Gang Li INTRODUCTION Statistical testing is one of the key tasks in quantitative tourism research. It is based on inferential statistics, which “based on probability theory and logic, are used to make inferences about the characteristics of a population from the characteristics of a random sample drawn from the population” (Grimm, 1993, p. 123). Statistical testing has been used for research purposes since the early 1700s (Huberty, 1993). Over the past 300 years of the development of statistical testing, four early twentieth-century statisticians, namely R.A. Fisher, J. Neyman, E.S. Pearson and K. Pearson, made the most significant contribution toward formalizing the concept (McLean and Ernest, 1998). Most of the current statistical testing techniques are still based on the same logic they developed. Specifically, the pioneers developed two different approaches to statistical testing: Fisher’s significance testing approach and Neyman–Pearson’s hypothesis testing approach. The former involves a single (null) hypothesis with a strength-of-evidence statistic p value, while the latter involves both a null hypothesis (H0) and an alternative hypothesis (H1) and specifies a fixed level of the probability at which the test statistic should be rejected. The significance testing approach does not typically consider the region of rejection or Type I and Type II errors (to be explained below), whilst the hypothesis testing approach does not consider the extent of support (i.e., the p value). The hypothesis testing approach is most often discussed in statistical methods for behavioral sciences. Although the two approaches are based on different philosophical beliefs, they share...