Research methods

The vast majority of the articles (78.8 percent: n - 41) were empirical in nature with 19.2 percent (n — 10) being reviews of the literature and 1.9 percent (n - 1) being conceptual. This strong emphasis on empirical approach aligned with Hinch et al.’s (2014) findings. Empirical studies included: case studies (43.9 percent: n = 18), questionnaire surveys (39.0 percent: n — 16), interviews (7.3 percent: n — 3), observations (2.4 percent: n — 1) and secondary analysis (2.4 percent: n - 1). Only two articles (Akiyoshi et al., 2014; Murata, 2010) employed mixed-method approaches. A similar dominance of case studies and questionnaire surveys was also reported in Hinch et al.’s review, such as articles focusing on one or a small number of sport events and sport tourism destinations.

However, the quality of the case studies in the systematic review of the literature in Japan was a major concern. Only a few of the case studies, such as Maeda (1998) met the demanding standards of a case study to ‘illustrate a problem; indicate a means for solving a problem; and/or shed light on needed research, clinical applications, or theoretical matters’ (American Psychological Association, 2009: 11). Similarly, in terms of the quantitative approaches, some researchers used advanced statistical techniques such as multiple group analysis (Sato et al., 2009) and structural equation modelling (Nishio et al., 2013; Park, 2015; Yumita & Harada, 2015) but most of the others (e.g., Emi, 2014; Watanabe et al., 2013) presented only descriptive statistics, such as means, or proportions, in their findings. Although Nogawa’s sport tourism publication series (e.g., Nogawa, 1992) also relied primarily on descriptive statistics, personal computers and statistical software were not easily accessible during this era. While this level of analysis has considerable value in terms of describing the state of sport tourism, it falls short of providing the explanatory power that more advanced statistical techniques can now provide. As Gibson (1998) argued two decades ago, discussions based on the results of sophisticated statistical analyses are needed for future quantitative research. In addition to this limitation, issues of measurement validity and reliability have tended to be overlooked. One example of this omission includes the failure to back-translate measurement instruments developed in another language (Brislin, 1970). These instruments should also be reviewed by experts to consider their applicability in a Japanese cultural context if they originated in a foreign cultural context. Conceptual and measurement invariance need to become important topics of discussion in cross-cultural sport tourism research. Lastly, the use of random, rather than convenience sampling is desired to enable a broader generalization of study results.

Qualitative research approaches, such as interviews or observation, were less common than quantitative approaches. As Ito et al. (2014) recommended, qualitative research helps researchers to provide the type of ‘thick description’ desired by anthropologists (Geertz, 1973) and cultural psychologists (Shweder, 1991). Qualitative research results in different types of insights than quantitative research with both types being necessary for the further development of sport tourism research (Gibson, 1998). However, even in the context of the qualitative studies that were found, only a handful of articles, including Oshimi et al. (2012) explicitly reported helpful details such as the nature of the interview questions (structured, semi-structured, unstructured) and the process used to develop the question lists. Similarly, the articles that purported to use observation-based methods failed to explain their procedures in detail (e.g., what behaviours the researchers tried to observe or how they recorded the data). While the advantages of triangulation are broadly recognized in research (Flick, 2017), the publications reviewed tended to focus on a single approach, failing to incorporate multiple theoretical approaches, methods and types of data.

Nishio (2011) was the only researcher who employed secondary data analysis as the primary method in his study. This is a missed opportunity as this research approach would open a major line of inquiry related to ‘big data’ collected by governments and companies. Cultural and institutional practices in Japan have long valued record keeping, which suggests that there are likely to be relevant databases that may allow sport tourism researchers to explore the big picture inclusive of spatial and temporal dimensions of sport tourism in Japan. Lastly, no examples were found in the review of experimental research, even though this approach has long been recommended for sport tourism research (Gibson, 1998). Although experimental methods have limitations in terms of the generalizability of the results, this approach has the major advantage of allowing researchers to measure the effect of the manipulation of an independent variable on dependent variables of interest (Vaske, 2008). It is, therefore, important for sport tourism researchers in Japan to move expand their repertoire beyond case studies and questionnaire surveys. Methodological innovation will contribute to the accumulation of sport tourism knowledge in Japan.