Contents lists available at ScienceDirect Journal of Outdoor Recreation and Tourism journal homepage: www.elsevier.com/locate/jort The tourism experience-led length of stay hypothesis Charles A. Adongo⁎, Frank Badu-Baiden1, Kwaku Adutwum Ayim Boakye Department of Hospitality and Tourism Management, University of Cape Coast, Ghana A R T I C L E I N F O Keywords: Backpackers Ghana Negative binomial Length of stay Tourism experience A B S T R A C T Even given the continuing interest of both academia and industry to understand what accounts for the length of stay (LOS) of tourists, explanatory variables have mostly been limited to socio-demographic and trip characteristics overlooking the influence of the holiday experience despite it being a major reason why tourism is embarked on. Departing from previous studies, this study proposed and tested an experience-led length of stay hypothesis employing a zero truncated negative binomial regression model. It was revealed that tourism experience significantly explains the variations in tourists’ LOS with self development, recreational engage- ments, hospitality, weather and sanitation identified as specific determinants. The findings also provide support for age, nationality, travel party size, budget, number of international trips, and risk taking behaviour as significant determinants of LOS. With these observations in mind, the study offers practical insights for sustaining tourists’ length of stay as well as propositions for future research on the tourism experience-led length of stay hypothesis. Management implications: Provision of positive memorable experiences in the domains of self development, recreation, hospitality, weather and aesthetics is one way to ensure that tourists stay longer. Specifically, need exits for destination management organisations and service providers to step up efforts in maintaining sanitation at the destination, especially at attraction sites and tourism-related premises. They can do this by ensuring regular cleaning, providing well-designated trash cans and disposing of sewage properly. In addition, a diverse of novel activities should be made available to tourists to sustain their interest and prolong their stay. Activity engagements that allow for skill acquisition such as volunteerism, cultural immersion (i.e. food bazaars) and co-creation of services are recommended. 1. Introduction Tourists’ length of stay (LOS) remains an important subject of interest for both academia and industry, given its positive association to tourism revenue (Thrane, 2012). Tourists who stay longer are believed to spend more than those who stay for shorter periods (Barros & Machado, 2010); meanwhile, studies (e.g. Barros & Machado, 2010; Ferrer-Rosell, Martínez-García, & Coenders, 2014) lament that tourism is increasingly characterised by reduced duration of stays. Campos-Soria, Inchausti-Sintes, and Eugenio-Martin (2015) consider reduced LOS as a phenomenon that adversely affects destina- tions, especially tourism-dependent economies. Measures must, there- fore, be in place to ensure that tourists stay longer at destinations. One way to do so is better understanding of the impact of tourism experiences on LOS (Martinez-Garcia & Raya, 2008). It is irrefutable that tourism is largely an experiential product, which is value of multiple on-site encounters revealed over time (Pine & Gilmore, 1999). Tourism experiences influence both current and future tourist behaviour (Barnes, Mattson, & Sorensen, 2016) and so, first it is reason- able to state that tourist on-site experiences can impact on-site behaviour such as LOS, during the trip. Research shows that a few days after the October 2002 Bali bombing, about 2000 tourists unexpectedly reduced their holiday LOS (Henderson, 2003). Second, if the statement that fulfilling tourism experiences predict pleasurable emotions, which in turn predicts positive behaviour is anything to go by (Barnes et al., 2016; Kim, 2014), it is sufficient to hypothesise that satisfactory tourism experiences positively relate to LOS while the reverse holds for adverse experiences. On the contrary, findings (see Hajibaba, Gretzel, Leisch, & Dolnicar, 2015) show that other tourists remain resilient during and after unfortunate events, suggesting that not all tourists will alter their travel itineraries due to unsatisfactory encounters at the destination. Obviously, this finding blurs the conventionally speculated direct relationship between holiday experi- ences and behavioural outcomes. Whether this is in reality the case or not, it needs to be substantiated. http://dx.doi.org/10.1016/j.jort.2017.02.003 Received 29 July 2016; Received in revised form 9 February 2017; Accepted 11 February 2017 ⁎ Corresponding author. 1 Current address: Department of Tourism Management, Takoradi Technical University, Ghana. E-mail addresses: charles.adongo@ucc.edu.gh (C.A. Adongo), baiden156@gmail.com (F. Badu-Baiden), kwaku.boakye@ucc.edu.gh (K.A.A. Boakye). Journal of Outdoor Recreation and Tourism 18 (2017) 65–74 2213-0780/ © 2017 Elsevier Ltd. All rights reserved. MARK http://www.sciencedirect.com/science/journal/22130780 http://www.elsevier.com/locate/jort http://dx.doi.org/10.1016/j.jort.2017.02.003 http://dx.doi.org/10.1016/j.jort.2017.02.003 http://dx.doi.org/10.1016/j.jort.2017.02.003 http://crossmark.crossref.org/dialog/?doi=10.1016/j.jort.2017.02.003&domain=pdf In spite of the growing number of studies (including Barros & Machado, 2010; Alén, Nicolau, Losada, & Dominguez, 2014; Santos, Ramos, & Rey-Maquieira, 2015; Thrane, 2016; Otoo, Agyeiwaah, Dayour, & Wireko-Gyebi, 2016) investigating the determinants of LOS, holiday experiences have rarely been considered as one possible determinant despite being a major reason why people embark on tourism. Explanatory variables used by past researches have included socio-economic and demographic variables (i.e sex, age, marital status, education and nationality), trip related characteristics (i.e party size, budget and past experience) and trip motivation (i.e adventure, relaxation, and cultural exchange). Remarkably, none of these variables is consistent in prediction of LOS across these studies, which is expected given the differences in study settings and encounters. This observation has been reinforced by Martinez-Garcia & Raya (2008) and Thrane (2016). Therefore, in line with suggestions by previous studies for further research to account for the residual variance in LOS, the current study proposes and tests the tourism experience-led LOS hypothesis using backpackers as empirical subjects. Nevertheless, the study provides no direct information about how experiences influence tourist adjustment of LOS while on-site. While this study is aware of the difficulty in defining a backpacker, past studies have often made recourse to Pearce (1990) conceptualisa- tion. With this frame of reference, backpackers are typically young tourists who travel alone or in small groups, prefer partly mediated and unstructured trips to unfamiliar destinations and seek contact with fellow backpackers and locals (Badu-Baiden, Boakye, & Otoo, 2016; Butler & Hannam, 2014; Hunter-Jones, Jeffs, & Smith, 2008). Backpackers are curious, motivated by novelty, seek active vacations and engage in adventurous activities (Loker-Murphy & Pearce, 1995). Moreover, they are described as tourists on limited budget and there- fore prefer budget accommodation facilities and usually carry their belongings in a backpack. Leiper (2010), however, highlights the increasing proportion of backpackers, who use packs-with-wheels, ‘wheelerist’, which among other reasons is attributed to their desire to avoid the physical strain of carrying a backpack. In regard to LOS, backpackers typically stay longer and are considered to spend more days at a destination than other travellers. Pearce (1990) in Australia observed that backpackers on the average stay 8 months. Similarly, Loker-Murphy and Pearce (1995) noted in their study that most backpackers stayed at least 6 months compared to other tourists whose stay was less than a month. Studies further note that for many backpackers the date of return is usually not decided in advance due a lot of time at their disposal and their quest for experiences furthest the ‘beaten track’ (Dayour, 2013a; Frew, McGeorge, Grant, & de Wildt, 2016). Reinforcing the "open return" nature of backpackers, Riley (1988) argues that backpackers are neither explorers nor drifters, but budget travellers. Budget travellers are “people wanting to extend their trips beyond that of a brief annual holiday and, therefore, faced with the necessity of living on a budget. The minimum period of travelling time required to qualify as a budget traveller is one year” (Loker-Murphy & Pearce, 1995: 825). Despite the extended LOS of backpackers, determinants of their LOS remain under-researched in the tourism literature. Moreover, backpackers are a growing market segment for developing destinations, which is explained by their high, direct and wide-spread spending in rural and deprived communities relative to their institutionalised counterparts (Dayour, Adongo, & Taale, 2016; Loker-Murphy & Pearce, 1995; Luo, Brown, & Huang, 2015). Finally, backpacking as an alternative form of tourism offers unique, diverse and meaningful experiences compared to other forms of tourism (Cohen, 1979, 2011), which makes backpackers ideal subjects when investigating the influ- ence of holiday experiences on LOS. Consequently, findings of this study are expected to add novel information to the tourism literature given the scarce information that exists on the impact of on-site experiences on LOS. To practice, insights are ultimately crucial since they can provide measures through which service providers can maximise benefits from tourists’ while they are still on-site. Maximisation of benefits from tourists while they are on- site is central since the association between satisfactory experiences and post- behaviour intention, including word of mouth publicity and revisit (Saayman, Krugell, & Saayman, 2016; Tung & Ritchie, 2011), is deemed hypothetical and actualisation cannot be guaranteed (McKercher & Tse, 2012). 2. Literature review 2.1. LOS LOS is conceptualised as the duration of an individual's leisure consumption and any other services or activities whose demand is prompted by virtue of visiting the destination. In the literature, LOS has been measured using various time cohorts. Most studies have looked at it in terms of short, medium and long duration of stay. However, what defines any of these time categories remains vague except for Otoo et al. (2016), who argue that periods between one month and a year yields better outcomes to both the tourists and the host. Another conceptualization of LOS was introduced by Thrane (2016) who considered it as two segments: “prefix” and “open” returners. For ‘prefix, the date of return is decided in advance, while with the ‘open’ returners it is partly or fully open. Cohen (1972) mass tourists and individual mass tourists may be typical cases of the former while explorers and drifters both akin to backpackers, best fit the latter. A majority of studies (see Box-Steffensmeier and Jones (2004), Alén et al. (2014)) have established that for many tourists the date is prefixed. It is extremely important to recognise, however, that depend- ing on the type of tourists and circumstances at the destination, it is possible that a tourist can cut short or extend his or her duration of stay. Data on tourism consumption is either collected before, during or after the holiday or at a combination of these stages. A review of the literature shows that the ‘during and after stages’ are common with LOS studies. The ‘during stage’ is also termed the on-site approach. This is where data is collected while respondents are still at the destination (see Otoo et al. (2016)). In other words, LOS is measured in-situ or concurrently with tourism consumption. Such studies usually ask of the intended LOS in view of the fact that the individual is yet to complete his or her holiday. In the situation that some or a majority of respondents are open returners, collecting data on LOS while the trip is ongoing risks measuring the variable inaccurately due to the possibility of extensions or reductions. The ‘after stage’ is also referred to as the ex-post facto approach where data is collected after respondents have exited (see Alén et al. (2014), Thrane (2016)) or are about to exit the destination (i.e. at the airport) (Aguiló, Rosselló, & Vila, 2017). Much as this stage is also liable to ex-post facto memory losses, it conceivably offers reliable information compared to the on-site method given that respondents may still have good recall ability and can provide accurate information about their holiday (Adongo, Anuga, & Dayour, 2015). As regards the mode of measurement, LOS has been measured from various competing perspectives of which a few are highlighted in this section. Some studies have considered it as a continuous variable (Otoo et al., 2016; Thrane, 2015), others as a count variable (Alegre, Mateo, & Pou, 2011; Brida, Meleddu, & Pulina, 2013; Salmasi, Celidoni, & Procidano, 2012) and others as a categorical outcome. In line with the varied modes of measurement, researchers have debated the appro- priate estimation technique for analysing LOS. Studies including Thrane and Farstad (2012), Lee, Alexander, & Kim (2014) and Scholtz, Kruker, & Saayman (2015) resorted to ordinary least squares regression [OLS] while Alegre et al. (2011), Salmasi et al. (2012) and Alén et al. (2014) employed poisson and negative binomial regressions [NBR]. The unsuitability of count models for estimating LOS in the tourism context has been argued given that they do not satisfy the condition of “how many times something has happened” within a C.A. Adongo et al. Journal of Outdoor Recreation and Tourism 18 (2017) 65–74 66 specified period of time (Thrane, 2015:1089). Further, in the light of simplicity and parsimony, Thrane Thrane (2015:1089) strongly re- commend OLS as the best estimator of variation in tourist LOS while acknowledging no variation in results regarding significance and sign of coefficients between the OLS and Poisson, NBR and survival models. Critiques of the OLS regression have raised concerns about the normal distribution of LOS as a dependent variable suitable for OLS estimation, likewise the ability of the technique to yield reliable results due to its simplicity (Martinez-Garcia & Raya, 2008; Alén et al., 2014). Survival models have also been utilised to analyse LOS based on its analogy with duration and survival research (Gokovali, Bahar, & Kozak, 2007; Martinez-Garcia & Raya, 2008). Nevertheless situations where travel consumption has already taken place, implying an absence of zeros, Box-Steffensmeier and Jones (2004) argue that it makes little sense to think of LOS as a positive random variable, which should be analysed using survival models. Other models such as the ordered logistic regression, one-way analysis of variance (ANOVA) and the independent samples t-test have received utility (see Ferrer-Rosell et al. (2014), Otoo et al. (2016)). Otoo et al. (2016) consider ANOVA/t-test as simple and easy to understand tools for analysing LOS, but these techniques have two main weaknesses. The first is their inability to estimate the joint effect of a set of independent variables on the dependent variable. The second is the use of series of a t-tests and ANOVAs, as is the case with Otoo et al. (2016) has the tendency to overestimate the significance of the results. Increasing different tests of mean differences can greatly inflate the likelihood of a false-positive outcome when evaluating statistical significance, which at a minimum a correction adjustment (e.g., Bonferroni) is recommended (Pallant, 2005). Bonferroni adjustment involves dividing the alpha value (i.e. 0.05) by the number of explanatory variables. Theoretically, this adjustment is necessitated as it provides a more stringent and reliable probability threshold to compare mean differences. However, this post-hoc method is equally criticized as overly stringent making it extremely difficult to detect significant results (Pallant, 2005). 2.2. Holiday experience Tourism is synonymously viewed as an experience commodity. In this sense, every encounter, be it behavioural or perceptual, cognitive or emotional, and expressed or implied, can be considered an experience (Oh, Fiore, & Jeoung, 2007). While defining what con- stitutes an experience has been difficult in the literature, recent studies (Adongo et al., 2015; Kim, 2014; Otoo & Amuquandoh, 2014) have observed that it is both an outcome of negative and positive encounters. A positive experience is considered as a favourable encounter (Kim, Ritchie, & McCormick, 2012). For instance, studies have found that tourism offers opportunities for personal development, social interac- tion, cultural exploration and enjoyment (Chen, Bao, & Huang, 2013; Larsen, Øgaard, & Brun, 2011). A study of Israeli backpackers by Uriely, Yonay, and Simchai (2002) revealed that some of them considered their trip as recreational, while others thought of it as developing knowledge. Negative experiences, on the other hand, represent encounters which are recalled unfavourably. This means that, though satisfying feelings is the primary motivation for partici- pating in tourism, one can unexpectedly experience dissatisfaction. In essence, the holiday experience is composed of mixed feelings from favourable and unfavourable encounters. Unique to backpackers is the fact that they stay longer and easily depart a destination for reasons unexplained in the literature (Loker-Murphy & Pearce, 1995). Such characteristics of backpackers are indications of complexity that is likely to surround the relationship between their on-site experiences and LOS. Like all other tourists, studies including Larsen et al. (2011) and Dayour (2013a) have highlighted backpackers as not homogeneous in their behaviour. On this score, it may be erroneous to assume that all tourists will behave the same way towards an encounter at a destina- tion. Cultural immersion is considered one of the motivations for back- packing. Backpackers strive to appreciate the authentic, tangible and intangible cultural attributes of a destination. These include food, dances, rites of passage and way people live at destination. Following from this, it is safe to propose that a significant relationship could exist between a destination's cultural heritage and LOS, with cultural attributes fulfilling expectations leading to longer stays and unattrac- tive cultural attributes causing truncation of LOS. The hospitality of the local people, for instance, can contribute positively to LOS, and the contrary may be true of hostility. Opportunity for social relations in a destination is argued to result in extended stays for backpackers (Jonas, Mansfeld, Paz, & Potasman, 2011). Social contact manifests itself in the form of an exchange through host-guest interactions. Encounters with destination attributes, such as climate, beaches and sanitation, have also been noted as potential determinants of LOS (Nicolau & Más, 2009; Prayag, 2012). Alén et al. (2014) understand that tourists who consider a destination's climate favourable tend to stay longer. Meanwhile, Akyeampong (2008) argues that visitors who consider temperatures in certain destinations to be harsh are more likely to reduce the duration of their stay. It implies that the climate of a destination can both facilitate and impede LOS (Hübner & Gössling, 2012). Aside entertainment and escapism, education and aesthetics were proposed by Pine and Gilmore (1999) in their idea of an experience economy. In both indoor and outdoor leisure context, aesthetics are deemed to impact the emotions of the consumer significantly. The engagement model highlights aesthetics of engagement as “the sub- ject's active, multisensory engagement in the environment, and the holistic, perceptual unity of the subject immersed in and continuous with their surroundings” (Todd, 2009: 161). Compared to the aes- thetics experience of everyday consumers, that of tourists’ goes beyond the visual aspects to include all senses. A study by Kirillova, Fu, Lehto, and Cai (2014) identified nine themes of aesthetic judgment at tourist destinations. These include scale, time, condition, sound, balance, diversity, novelty, shape, and uniqueness. Kirillova et al. (2014) indicate that the uniqueness of a destination's attributes can exert both short and long term effects on tourist behaviour. This suggestion entails that the aesthetic properties of a destination are influential not only in evaluating the entire consumptive experience but in stimulating consumption decisions including duration of stay. Satisfactory experiences are the primary motives for participating in tourism, but tourists can unexpectedly realise negative moments (Kim, 2014). These include economic crises, crime, pollution, epi- demics and terrorist attacks (Boakye, 2010; Holcomb & Pizam, 2006; Ryan, 1991). Studies on service quality have also noted various areas of tourist dissatisfaction ranging from poor services to purposeful, financially- driven service sabotage by service providers (Harris & Ogbonna, 2002). Behavioural theorists have considered the behaviour of service providers as impacting on customers experience negatively with two main terminologies, namely ‘counterproductive behaviour and misbe- haviour’ (see Harris (2012)). Commonly reported experiences of tourists include dupes, fraud, deceit and theft (Adongo et al., 2015; Harris, 2012; Adam & Adongo, 2016). Such outcomes subsequently lead to anger and frustration (Kim, 2014). Mixed findings are presented in the literature on the possible effect of adverse encounters on LOS. Some studies (Boakye, 2010; Campos- Soria et al., 2015) opine that adverse encounters at a destination can hold devastating consequences at the destination and nearby destina- tions. These include trip cancelations, shortened stays and abrupt drops in demand. By late 2004, though Bali's tourist arrivals revamped, visitors were staying for shorter days and spending less (Gurtner, 2016). Law (2006) concludes that the majority of tourists, particularly Asian tourists, change their travel itineraries when faced with a risky C.A. Adongo et al. Journal of Outdoor Recreation and Tourism 18 (2017) 65–74 67 circumstance at a destination. However, others are of the view that the effect of such events is infinitesimal. For Hall (2010), unless a negative encounter persists for a substantial length of time, it is exceedingly unlikely to have a negative impact on the visited destination. Similarly, Hajibaba et al. (2015) in their theorisation of crisis resistant tourists argue that some category of tourists remain resilient in the face of unexpected events. Therefore, not all tourists will alter their stay in the face of unfortunate events. In a more specific context, given back- packers’ quest for novelty, it is difficult to situate their behavioural outcomes amid negative encounters. Closely related to negative events is the fear attached to it. Fear is a psychological state that almost every human will encounter at one point of his or her life. It is a state of uncertainty, worry, anxiety and distress caused by the awareness or expectation of danger or pain. It is common when an individual is on the verge of or encounters the unknown such as, an unfamiliar destination. In prospect theory, Kahneman and Tversky (1979) theorise that fear results when per- ceived losses associated with an event outweigh the benefits (Kahneman & Tversky, 1979). Fear can also be driven from several other sources such as one's own experience, friends and relatives and mass media. The literature indicates a clear gap between development of fear and its actual occurrence. Behavioural economists have since time immemorial thought that the fear of an encounter is much larger than its occurrence. In other words, Stanko (2000) in his paradox of fear argues that most things people think will occur never do occur. For the tourists, Boakye (2012) argues that fear correlates negatively with tourism demand. Money and Crotts (2003) established that tourists most likely to avoid uncertainty were also probable to undertake shorter stays. For backpackers, whereas there have been situations where the thrill of danger is part of the experience (Hunter-Jones et al., 2008) challenging the dominant idea that fear is a negative psycholo- gical construct, Adam (2015) pointed out some perceived risks of backpackers. 3. Materials and methods The context for the study is Ghana. Ghana is a West African country bounded by Togo, Burkina Faso, and Ivory Coast, all of which are Francophone countries. This makes the destination preferred by tourists from Anglophone countries, particularly gap year students. It has been noted as one of the emerging destinations that receives a significant share of the backpacker market mainly due to its incipient stage of tourism development with many opportunities to explore the unknown. These include but not limited to ecological, cultural, climatic and coastal features. It also hosts a significant proportion of budget accommodation and traditional catering establishments. According to Dayour (2013a), backpackers’ arrivals to Ghana have more than doubled over the last four years. This is especially so for the Cape Coast- Elmina Conurbation (Dayour, 2013a, 2013b). It is therefore not out of place that it is one of the core objectives of the 2013–2027 National Tourism Development Plan of Ghana to identify ways of maximising benefits from backpacker tourism (Ministry of Tourism, 2014). Taking cue from Chung, Kyle, Petrick, and Absher (2011) and Dayour et al. (2016), a two-phase approach was used to collect data for this study. The data collection period spanned between July and October 2015. The first phase involved contacting backpackers at budget accommodation facilities. Budget accommodation facilities have been used by most studies as an effective means to locate backpackers (Cohen, 2011; Dayour, 2013a; Luo et al., 2015). The major drawback of this approach is the assumption that all backpackers prefer budget facilities. This overlooks the growing diversity of back- packing such as flashpackers who are deemed to have higher income and can patronise upscale accommodation facilities (Hannam & Diekmann, 2010). Nevertheless, the justification for the reliance on the budget accommodation approach is the observation by Butler and Hannam (2014) that though flashpackers have the wherewithal to patronise upscale hotels they still prefer budget hotels because they are considered to be more sociable. Subsequently, 25 budget accommodation facilities comprising hostels, guesthouses and homestays were selected from the accommo- dation facilities register of the Ghana Tourism Authority (Ghana Tourism Authority, 2014) in Central Region for the study. Respondents were conveniently selected during check-out at the front desk of the accommodation facilities. In identifying backpackers, the study adopted two approaches used by Hunter-Jones et al. (2008) and Adam (2015). One should have visited entirely for leisure and must have identified himself or herself as a backpacker. The rationale for employing the ‘self-identification’ method was to ensure that valid and reliable respondents are sampled. This is perhaps superior to relying on the seemingly infinite number of criteria proposed in the literature, such as the age based approach that may result in unreliable samples. For example, an individual may be young and yet belong to another typology of tourists than backpackers. Aside the filter questions, background information such as sex, level of education, intended level of stay, and email addresses were collected using questionnaires. In all, five hundred and fifty (550) backpackers with valid particulars consented to participate in the study at the first stage. Based on the intended LOS of each respondent sampled in the first stage, the fieldwork (second stage) consisted of sending questionnaires to each respondent, through e-mail, five days after the intended departure date. This was to guarantee that potential respondents returned to their places of origin and were able to provide accurate information on their experiences and LOS. The short fallow time period between departure and receiving the questionnaire guarded against any likely ex-post-facto memory loss. The main questionnaire, worded in English, was emailed to the respondents and 400 (73%) usable cases were received. The questionnaire was supplied only in English because almost all respondents who were contacted during the first stage of the study demonstrated mastery in both oral and written English. Furthermore, evidence exists that a majority of backpackers who visit Ghana are fluent in English (Adam, 2015). The instrument was divided into two (2) sections: The first section sought information on backpackers’ experiences through the use of open-ended questions. Respondents were asked to freely recall their experiences in Ghana. While there abound a number of standard scales for measuring tourism experiences, the decision not to use any of those scales but to rely on the self-reporting technique was based on the reasoning that people's experiences may differ per setting and context. More importantly, usage of already existing scales largely predisposes respondents to issues already identified in the literature (Otoo & Amuquandoh, 2014), hence a limitation to generating new informa- tion. The second section covered their length of stay and other socio- demographic characteristics, trip characteristics and trip motivation. All issues were investigated at the wider destination context. STATA, version 14, software was used to process the data. Analysis of the self narrated holiday experiences began with manual coding of every single commentary in the received questionnaires. With recourse to Kim (2014) memorable attributes of a destination, a coding method which involved compilation of responses and assignment of codes to responses was adopted to capture the responses (See Table 1). First, caution was taken to ensure that the codes mirrored all the respon- dents’ verbalisations. Second, the verbalisations were coded indepen- dently by the three authors and additional independent judge. Inter- judge reliability was 83 percent, exceeding the accepted threshold of 80 percent (Latham & Saari, 1984; Patterson, Brady, & McColl-Kennedy, 2016). Seven themes emerged namely; personal development, local culture, recreational activities, aesthetics, infrastructure, rip off and fear. Subsequently, the NBR truncated at zero was used to estimate the influence of the holiday experience on backpackers’ LOS. Travel to a destination is an indication that there is no longer a zero outcome in C.A. Adongo et al. Journal of Outdoor Recreation and Tourism 18 (2017) 65–74 68 the travel decision, which makes the conventional NBR model inap- plicable in this context. The alternative to this is to employ a zero truncated NBR model (Alén et al., 2014). Further, the suitability of the zero truncated NBR model for this particular data set was first based on the mode of measurement of the dependent variable. LOS was measured as number of overnight stays over a period of at least a day and a maximum of 366 days, hence a count variable. Aside purpose of visit, the period of stay variable is well agreed on in the tourism literature as a measure of the temporariness of one's movement qualifying him or her as a tourist. It is important to note however that if LOS is measured in time (hours, minutes and seconds), it cannot be considered a count variable. Second, we are convinced that the manner with which LOS is measured in this study not only makes it a positive integer but equates with “how often something has happened during a specific period of time” (Thrane, 2015:1089). Finally, the NBR model has been employed by previous studies (see Alén et al. (2014), Losada et al. (2016)) and yielded reliable and valid results. Poisson regression is often the first to be considered for fitting count data, but the NBR was found appropriate because the Pearson likelihood-ratio test indicated that the mean and variance of the data were not equal implying over dispersion with respect to the Poisson distribution parallel to what appears to be the case in general (Cameron & Trivedi, 2005). This result is the refutation of the conditional assumption of equal mean and variance required for Poisson models (Ajiferuke & Famoye, 2015). The function for the negative binomial distribution is given as ⎛ ⎝⎜ ⎞ ⎠⎟P Y y a y y θ y( − ) = 1/ + − 1 (1 − ) , = 0, 1, 2, 3...a θ 1/ y (1) The mean and variance of the above model are respectively specified as μ θ a θ= /[ (1 − )] and σ θ a θ= /[ (1 − )2 2] From this, it can be seen that the NBD is over-dispersed, that is the variance exceeds the mean. Assuming that the mean backpackers’ length of stay depends on some predictor variables xi, then we can write μ x θ a θ( ) = /[ (1 − )]1 and this gives θ aμ x aμ x= ( )/[1 + ( )]1 1 . Thus, the NBR can be written as ⎛ ⎝⎜ ⎞ ⎠⎟ ⎛ ⎝⎜ ⎞ ⎠⎟ ⎛ ⎝⎜ ⎞ ⎠⎟P Y y x a y y aμ x aμ x aμ x aμ x y( = / ) = 1/ + − 1 ( ) 1 + ( ) ( ) 1 + ( ) , = 0, 1, 2, 3...i i i i yi a i 1 1 1 1 1/ (2) 3.1. Sample characteristics The sample characteristics reported in Table 2 suggest that the majority of respondents were females (61.30%). The proportion of respondents who were unmarried was more than those who were married, constituting 87.20 percent and 12.80 percent respectively. The average age of the respondents was 25 years. Most respondents (70.00%) professed Christianity as their religion. With respect to level of education, more than half of the respondents (55.60%) had attained university degrees. About 60 percent were Europeans and the smallest percentage being African (10.00%). On average, respondents stayed 24 days and a maximum of 240 days. The minimum budget was US$5000 and the maximum US$32000. The average number of international trips made by a respondent in the past was eight (8) and a majority visited Ghana without a pre-planned itinerary (86.6%) whereas 13 percent did. As regards trip motivation, about 29 percent visited for cultural reasons, 24.1 percent did for ecological reasons and 17.1 percent did for relaxation. Table 1 Codebook of holiday experiences of sampled backpackers in Ghana. Source: Fieldwork, 2015. Outcomes Definition Personal development “an eye opening expedition”, “so much to learn”, “happy being there”, “had a lot to share”, “learnt how to greet in twi [etisen]” etc Local culture “friendly people” “fun to be with the kids” “great culture”, “different culture”, authentic foods”, “different traditions”, fascinating culture” “sweet people” etc Recreational activities “lots of varied things to see and do”, “helping others during free time is accessible” “there is enough to do”, “great farming experience” etc Aesthetics “good landscape”, “natural beaches”, “nice weather” “new climatic experience” “garbage everywhere” “dirty beaches”, “poor sanitation” etc Poor infrastructure “bad roads”, “constant blackouts”, “no flashing toilets”, “lack of wifi” “no destination apps”, “difficult getting transport in rural communities”, “unreliable water supply” etc Rip- off “people think that you are rich so they screw you”, “a lot of begging along the coast” “pricing discriminated” “scam” etc Fear “unable to walk without a guide”, “people are pushy”, “scary places” “religious extremism” Table 2 Socio-demographic and trip characteristics of backpackers in Ghana. Source: Fieldwork, 2015. Characteristics N Percent Mean Min Max Sex Male 157 39.00 Female 243 61.00 Marital Status Married 51 12.80 Unmarried 349 87.20 Age 400 25.00 19 56 Highest level of education Senior High School 109 27.30 University 222 55.60 Post Graduate 69 17.10 Nationality Europe 240 60.00 America 76 19.00 Australia 44 11.00 Africa 40 10.00 Backpacking history to Ghana First timers 346 86.60 Repeat visitors 54 13.40 Party size 1 70 Travel budget ($) 338 1368.13 5000.00 32,000.00 Number of past international trips 385 11.21 1 96 Length of stay 400 24.00 7 240 Travel itinerary (N=393) On itinerary 52 13.20 Not on itinerary 341 86.80 Risk personality Risk averse 76 19.00 Risk taker 324 81.00 Trip motivation ( N=1200*) Ecological 289 24.10 Relaxation 205 17.10 Culture 493 41.10 Learning 213 17.70 C.A. Adongo et al. Journal of Outdoor Recreation and Tourism 18 (2017) 65–74 69 4. Results 4.1. Independent variables Seven (7) themes of holiday experience were recalled by the respondents. These included personal development, local culture, recreational activities, aesthetics, poor infrastructure, rip off and fear. The most memorable were the cultural attributes of the destination (27.74%). Statements to that effect included novelty of the cultural practices (46.67%), friendliness of locals (28.94%), and uniqueness of the destination's local cuisines (24.39%). Some satisfactory experiences were also recalled about the aesthetics of the destination (27.70%) as shown in Table 3. The places visited were described as unique and novel (17.10%). They labelled the tropical weather as congenial (10.91%) and the beaches as natural (8.79%). Others regarded their journey to the destination as a means to personal development (11.68%). They described the holiday as an image enhancer, an avenue to skill acquisition and to boost their future career (Table 2). Adverse feelings about the destination were also reported. The majority (63.19%) lamented about insanitary conditions, particularly improper waste disposal and littering. In addition, senti- ments of rip-off and fear were expressed (Table 3). About 51.48 percent noted that they had been over-charged at some point, of which transport service providers (largely taxi drivers) and food and beverage workers were the main culprits, representing 34.10 percent and 12.30 percent respectively. 4.2. Econometric results Table 4 presents the results of the hierarchical NBR. Following from the literature review, this first sought to isolate the influence of socio- demographics and trip characteristics on LOS in Model 1. This was followed by the main explanatory variable, holiday experience, in Model 2. The last model (Model 3) tested a joint effect of the variables in Model 1 and 2. The first model significantly explained 13 percent of the variance in LOS (R2 =0.130; LR =51.30; p < 0.05). Controlling for socio-demographics, trip characteristics and trip motivation, though they were not likely to confound the effects of holiday experience, improved the precision of the study's estimation. The second model was also significant and explained 15 percent of the variance in LOS. The final model showed that both sets of predictors accounted for about 27 percent of the variance in LOS, but the coefficients give indications of varied effects (Table 4). Notwithstanding the small variation in explanatory power, statistical evidence exists that holiday experience independently predicts LOS significantly, but an improved model is obtained when analysed jointly with socio-demographic and trip characteristics. Due to different estimation techniques employed by past studies such as OLS (Thrane & Farstad, 2012; Thrane, 2012), Poisson and NBR (Alén et al., 2014; Prebensen, Kim, & Uysal, 2015) and ordered logit models, Thrane (2015) argues convincingly that there is no conventional benchmark for what might pass as acceptable model fit. However, comparing the resulting R-square in this study to those of previous studies that used NBR, it is evident that the output in the present case is satisfactory. Alén et al. (2014), for instance, yielded an R-square value of 0.248. For statistical efficiency and parsimony, the third model is preferred and thus the individual effects of the main variables of interest, experiences, are turned to. It is observed that friendliness of the local people is a significant determinant of LOS with the coefficient suggesting that tourists who consider the destination to be hospitable are by a factor of 2.364 more likely to stay longer. A similar observation is made for the cultural practices of the destination. Cultural attributes that were considered novel increased LOS by a factor of 1.248. The study's results further give an indication that LOS is positively influenced by congenial weather (Z =3.544; p < 0.01) such that respondents who considered the destination's weather as congenial tended to stay longer. Moreover, varied activity engagement appeared as one of the most important experiential attributes significantly influencing LOS by a factor of 1.248. The indication is that an increasing number of activities corresponded with increasing likelihood to extend stay. Nevertheless, the study notes that insanitary conditions significantly shortened LOS (Z=2.15; p < 0.01). This suggests that tourists who are dissatisfied with the sanitary conditions of a destination have higher chances of short- ening their stay compared to those who are satisfied with sanitary conditions. A similar relationship was established between ‘being overcharged’ and duration of stay (see Table 4). As regards the control variables, evidence is found for age, nationality, travel party size, travel budget, number of past international trips and risk taking behaviour as significant determinants of LOS. 5. Discussion Considerable discourse exists stating that holiday experiences are relevant for all segments of tourists. Therefore, it is extremely important to understand how holiday experiences impact LOS. The study filtered seven dimensions of holiday experiences, which three included local culture, aesthetics and infrastructure. These are con- sistent with three of the 10-dimensional memorable attributes of a destination proposed by Kim (2014). In contrast, the dimensions: recreational activities, personal development, rip-off and fear are unique to this study. The range of experiences reported also confirmed the notion that holiday experience is multifaceted with positive and Table 3 Descriptive statistics of holiday experiences of backpackers in Ghana. Source: Fieldwork, 2015. Typology Variables included N Percent (%) Total (%) Personal development (N = 259) Enhanced image 158 61.00 11.68 Enhanced future career 84 32.43 Acquired skills 17 6.56 Local culture (N = 615) Unique cuisines 150 24.39 Friendly people 178 28.94 27.74Novel cultural practices 287 46.67 Recreational activities (N = 259) Interesting local working life/volunteer 105 40.54 11.68 A lot of varied activities to participate in 154 59.46 Aesthetics (N = 614) Insanitary condition 388 63.19 Unique places to visit 105 17.10 27.70 Congenial weather 67 10.91 Natural beaches 54 8.79 Infrastructure (N = 138) Unreliable electricity supply 62 44.93 Transport problems 63 45.65 6.22 Unreliable water supply 13 9.42 Rip-off (N = 270) Swindle 69 25.56 Perceived overcharging 139 51.48 12.18 Corrupt officials 62 22.96 Fear (n = 62) Fear of crime 16 25.81 Fear of diseases 30 48.39 2.80 Religious extremism 16 25.81 Total 2217a 100.0 a Frequency exceeds 400 because of multiple responses C.A. Adongo et al. Journal of Outdoor Recreation and Tourism 18 (2017) 65–74 70 Table 4 Influence of backpackers’ socio-demographics and trip characteristics and holiday experience on their LOS in Ghana. Source: Fieldwork, 2015. Model 1 Model 2 Model 3 Coef (SEa) Z Coef (SEa) Z Coef (SEa) Z Socio-demographics characteristics Sex Male (ref) Female 0.328 (0.135) 2.43*** 0.273(0.263) 1.04 Marital status Unmarried (ref) Married 0.139 (0.224) 0.62 0.190 ( 0.421) 0.45 Age −0.045 (0.013) −3.35** −0.057 (0.021) −2.63** Education High school (ref) Bachelor 0.067(0.144) 0.47 0.419 (0.265) 1.58 Post Graduate −0.140(0.296) −0.47 −0.232(0.460) −0.51 Nationality (ref. Africa) Africa (ref) Europe 0.2443(0.167) 1.67 0.199(0.123) 1.78 North America 0.243 (0.167) 1.49 −0.009(0.259) −0.03 Australia −184 (0.168) −1.09 −0.578 (0.296) −1.95* Trip-related characteristics Backpacking history to Ghana Repeat visitor (ref) First-timer Party size 0.016( 0.009) 1.83 0.045( 0.016) 2.71** Travel budget 0.732 (0.103) 3.12** 1.562 (0.123) 1.96* Number of past international trips 0.074 (0.068) 1.09 0.2093 (0.089) 2.33* Travel itinerary On itinerary (ref) Not on itinerary 0.237( 1.294) 1.83 0.199 (0.234) 0.85 Risk personality Risk averse (ref) Risk taker 0.819 (0.387) 2.12* 1.552 (0.818) 1.90* Trip motivation Local culture (ref) Nature −1.83 (0.169) −1.08 0.090(0.255) 0.36 Relaxation −0.054 (0.165) −0.33 0.006 (0.237) 0.03 Learning 0.065(0.273) 0.24 −0.797(0.926) −0.86 Holiday experiences Personal development Enhanced future career (ref) Enhanced image −0.184 (0.242) −0.76 0.184(0.242) −0.76 Acquired skills 0.04 9 (0.247) 0.20 (0.088) 2.40** Culture Unique cuisines (ref) Friendly people 0.302 (0.819) 0.37 2.364(0.506) 4.02** Novel cultural practices 0.503 (0.263) 1.91 1.248 (0.109) 2.53** Recreational activities Interesting local working life/volunteerism (ref) Varied activities to participate in 1.020(0.206) 4.05** 1.248(0.109) 3.05** Aesthetics Unique places to visit (ref) Insanitary conditions −0.390(0.262) −1.49 −1.299(0.158) −2.15* Congenial weather 0.431(0.279) 1.54 1.384(0.127) 3.54** Natural beaches 1.023 (0.221) 1.97* 1.523(0.211) 3.03** Infrastructure Unreliable water supply (ref) Unreliable electricity supply 0.215(0.143) 0.73 0.915(0.132) 0.61 Transport problems 0.028 (0.185) 0.15 0.094(0.098) 0.58 Rip-off Corrupt officials (ref) Swindle 0.196 (0.227) 0.86 0.196(0.227) 0.86 Perceived overcharging −0.390(0.203) −1.91 −0.390 (0.203) −1.91* Fear Religious extremism (ref) Fear of diseases −0.053 (0.187) −0.28 −0.053 (0.187) 0.28 Fear of crime 0.166 (0.250) 0.65 0.166(0.255) 0.65 Constant 1.644(0.496) 3.31** 1.434(0.123) 4.31** 2.452(0.091) 4.99** (continued on next page) C.A. Adongo et al. Journal of Outdoor Recreation and Tourism 18 (2017) 65–74 71 negative outcomes signalling that a holistic two-sided conceptualisa- tion is sufficient to capture the variance. The mixed nature of experiences should act as a further caution to the one-sided assumption and subsequent measurement of holiday experiences as either positive or negative adopted by some studies (Kim et al., 2012). The experiences reported by respondents appeared to be con- structed around the tangible and intangible attributes of the destina- tion as well as the respondents’ own intrinsic needs. Accordingly, these experiences significantly influenced LOS. For instance, positive re- marks about the appealing attributes of the destination, which included the novelty of the cultural practices, hospitality of the people and congeniality of the weather, positively impacted LOS. An alternative interpretation for this observation is the significant association found to exist between trip motivation and length of stay at the destination (Alén et al., 2014). Similarly, trip motivation provided some moderate explanations to the main explanatory variable implying that tourists’ experiences are not only the outcome of their actual encounters, but also the result of their motivations and expectations. However, con- trolling for motivation in the NBR model, no significant influence was observed on LOS. While it is conceded that stimuli, such as on-site experiences, can blur the recollection of motivations leading to con- founding effects, it is established that experiences predict LOS better than trip motivations. A case in point is the fact that a tourist in search of personal development was more likely to stay longer as well as to actually experience personal development, but the fact remains that it is the in-situ experience of personal development that predicts the LOS. The argument in the literature is that positive experiences lead to pleasant outcomes whereas negative experiences lead to unpleasant outcomes. While this study has to some extent substantiated this supposition, the findings suggest that specific holiday experiences impact LOS differently. This suggests that while on-site experiences on the overall may significantly influence LOS, it is not so in terms of magnitude and direction with respect to specific experiences, and hence different behavioural outcomes can be expected. Aesthetic experiences played a key role in influencing LOS, especially congeni- ality of the destination's weather and naturalness of beaches, leading to longer stays. Akyeampong (2008) avers that in this age of environ- mentalism, tourists from developed countries opt more for products and services in unique and natural settings. This idea mirrors Ryan et al.'s. (2003) description of backpackers as avid appreciators of nature and exclusiveness. Tourists attribute the value of uniqueness to a place when its features are exceptional, while novelty denotes originality of the features (Ryan et al., 2003). Affection for nature engages most of the human senses resulting in higher order cognitions and emotions which leads to intimate destination attachment. For Pröbstl-Haider, Haider, Wirth, and Beardmore (2015), good climate and serene beaches enhance the recreational activities of tourists. Contrary to Alén et al. (2014) who observed no significant relationship between cleanliness of destination and LOS, this study notes that insanitary surroundings significantly shortened LOS. Unclean surroundings affect the allure of destinations thereby making it physically unattractive to visitors who would have wished to stay much longer (Kim, 2014). The fact that most of the diseases resulting from tourism are usually linked to poor sanitation, inadequacies of water supply and sewage disposal (Henderson, 2007; von Schirnding, 2002) coupled with the discomfort such conditions can pose obviously makes reduced LOS an avoidance strategy one would opt for. Similar to studies (including Crouch, 2011; Ferrer-Rosell et al., 2014) that have acknowledged the need for diversification of activities to enhance a destination's competiveness and its delivery of memorable experiences, the current study highlights its role in sustaining visitors’ stay. It was observed that respondents who engaged in varied activities significantly stayed longer. Variety they say is the spice of life, which gives an indication that when there are varied things to see and do, the odds of staying longer is higher. Evidence indicates that participating in diverse activities impacts holiday satisfaction positively (Kim et al., 2012). Cultural immersion has widely been documented as one of the most sought-after experiences by tourists. On that score, it was only natural that novel cultural practices and hospitality of the local people positively influenced LOS. The hospitality of host residents towards a visitor can be interpreted as a mark of acceptance and social solidarity, and so chances are that backpackers in need of fellowship with local people would stay in the destination longer. Nonetheless, among mainstream tourists, Menezes, Moniz, and Vieira (2008) observed that those who attached priority to cultural heritage tended to have shorter stays. The conception of the holiday as a means to personal development hinged on three issues: recognition, skill development and employ- ability (Chen et al., 2013). Skill development has to do with progressive mastery of skill(s), whereas employability represents an increment of one's chances of being offered a job, and recognition deals with the psychological satisfaction derived from being dignified by others (Arnould & Price, 1993; Tung & Ritchie, 2011). Whereas skill acquisition directly and positively influenced LOS, the reverse was noted for enhanced image, though not significantly. Time is an important indicator of skill mastery, which obviously would manifest in longer stays. On the other hand, Dann (1977) avows that holiday returnees stand a greater chance of ego enhancement by virtue of recounting their holiday experiences to others. A plausible reason is that a strong desire for recognition could mean that one would have to return home quickly to be hailed by peers and relevant others. It is further observed that overcharging adversely impacts LOS. Explanations for this outcome lie in the fact that tourism is financed by discretionary income and increasing length of stay requires an addi- tional budget (Thrane, 2016). In such instances, depending on the amount of money lost as a result of overcharging, the already constrained budget of the tourist can be worsened leaving him or her with unsustainable finances that can make him or her to reduce the planned duration of stay. This observation indicates that the exploita- tive tendency of service providers when noticed by clients can adversely affect purchases and LOS at the destination. Finally, the NBR model provides support for the controlled vari- ables, namely age, nationality, travel party size, travel budget, number of international trips and risk taking behaviour as significant determi- nants of LOS. A significant negative relationship between risk taking and non-cancellation of trips despite adverse events has been estab- lished by Hajibaba et al. (2015). They aver that high risk propensity is considered a psychological driver of exploration and endurance to travel despite adverse experiences on-site. Risk taking backpackers Table 4 (continued) Model 1 Model 2 Model 3 Coef (SEa) Z Coef (SEa) Z Coef (SEa) Z LR statistic 51.30 56.32 86.65 R2 0.1307* 0.1823* 0.2702** ΔR2 0.1307 0.0516 0.0879 ** 99% significance level and Ramsey’s p-value of 0.121. * 95% significance level. a The values in parenthesis are robust standard errors (SE). C.A. Adongo et al. Journal of Outdoor Recreation and Tourism 18 (2017) 65–74 72 staying longer than those who are risk averse is in part an important explanation as to why some tourists would abruptly shorten their holiday during adverse encounters. On the other hand, it also reflects the assertion that the thrill that comes with risk can be a motivating factor for travel (Fuchs & Reichel, 2011). Despite the fact that previous studies have highlighted the relevance of age, nationality, travel party size and travel budget to LOS, little attention has been devoted to past number of international trips and risk taking behaviour. The number of international trips taken in the past signals globetrotting—travelling widely and regularly all over the world, including intra destination suggesting a desire to explore and hence a positive association with LOS. Contrary to studies that point to a positive relationship between aging and LOS at destinations (see Allen et al. (2014)), the opposite was established in the context of backpacking. 6. Conclusion and implications The study sought to find empirical evidence in support of holiday experience as one of the explanatory factors of LOS or otherwise. The findings confirm that backpackers’ holiday experience is a significant determinant of their LOS. Specifically, skills acquisition, unique and natural attributes of the destination, varied recreational activities and hospitality of the local people lead to longer stays. Essentially, these experiences matter to backpackers. On the contrary, image enhancement, insanitary conditions, and overcharging significantly reduced LOS. This means that the backpackers’ LOS is a complex one and does not depend on a single experiential issue. More importantly, there is compelling evidence to conclude that reduction in LOS due to negative experiences at the destination is an expression of dissatisfaction and its directed behaviour. Theoretically, the current study has confirmed the importance of socio-demographics and trip characteristics in explaining LOS, but distinctively it contributes to the extant tourism literature in two main ways. First, it provides empirical evidence on experience as a determi- nant of LOS. This draws not only the attention of policy makers to holiday experience as a promising segmentation variable to LOS but scholars on the need to incorporate behavioural data in a bid to understand LOS. In keeping with this study, future attempts to model the relationship between tourists’ experiences and their LOS should endeavor to disaggregate and capture the two-sided (negative and positive) components of holiday experiences. In addition, the NBR employed in this study validated the assertion that behavioural data are often characterised by over dispersion. Thus, this study advances two issues. First, similar to Alén et al. (2014), the study concludes that NBR is more fitting than the Poisson model in analysing the association between experiences and LOS, as it guards against the likely bias of the estimates. Second, usage of econometric techniques that can handle heterogeneity in tourists’ behaviour while predicting their LOS is critical to providing nuanced and valid findings. The study contributes to practice by addressing the determinants of LOS of backpackers, a growing market segment which has seldom being given attention. Regarding the study's practical implications, one could not have agreed more with the conclusion that a majority of tourists are pre-fixed returnees (Thrane, 2016), but since backpackers travel on flexible itineraries it is highly possible that satisfactory on-site experiences will lead to extended stays. Efforts at minimising the negative experiences of tourists increase revenue for destinations by lengthening LOS. With this thinking, strategies aimed at persuading backpackers to stay longer at a destination are certainly worthwhile, contrary to Thrane (2016) who suggested that such efforts are a waste of resources. This study's results further hold implications for destination planning, branding and management. It is suggested that in adding aesthetic values to attractions such as beaches, cultural heritage and historical relics, it is needless for DMOs to incur extra costs to overly develop and ‘westernize’ their attractions. Providing basic infrastructure such as good roads, electricity and sanitation will suffice. With specific mention to Ghana's beaches, Akyeampong (2008:11) highlights that “whereas there is the need to add a little more value to the beaches, concrete, mortar and steel are hardly the ingredients needed… improving accessibility and sanitary conditions as well as safety should be enough”. Furthermore, given the continuous desire among people for self-improvement (Chen et al., 2013), backpacking as a complimentary tool to formal education for self- development could be emphasised in destination marketing campaigns to promote extended stays among inbound tourists. Acknowledging heterogeneity in backpacking, the authors regard a long-staying backpacker as predominantly someone who is a young globetrotter, travels in the company of others and is risk averse, celebrates skill acquisition, novelty, varied recreational activities and is sanitation conscious. This is germane for market segmentation. Much as the study provides measures to sustain backpackers’ LOS, it recognised that backpackers’ stay in local communities is not sacro- sanct as it can lead to significant adverse impacts if not properly managed. In that regard, DMOs should thoroughly orient host resi- dents on the pros and cons of having backpackers in their communities. Despite the relevant contribution of the study to theory and practice, more research is called for to validate the findings. Its cross-sectional nature did not make it possible for time variation evidence, thus a longitudinal study will be commendable. The analysis could not explicitly account for how backpackers adjusted their LOS along positive or negative experiences or a combination. This could be another interesting issue worth addressing in which case duration models would be more appropriate. Finally, it is reasonable to think that tourist experiences may influence their LOS in the next trip to the same destination and thus would be of value when researched. Acknowledgements We are immensely grateful to Eva Adomaa Kyereboah of Kansas State University and Ryan Carty for their invaluable comments during the preparation of the manuscript. We would also like to thank the journal editors and anonymous reviewers for their constructive comments and suggestions. However, the content is the responsibility of the authors and do not necessarily reflect the views of the above mentioned. References Adam, I. (2015). Backpackers' risk perceptions and risk reduction strategies in Ghana. Tourism Management, 49, 99–108. Adam, I., & Adongo, C. A. (2016). Do backpackers suffer crime? An empirical investigation of crime perpetrated against backpackers in Ghana. Journal of Hospitality and Tourism Management, 27, 60–67. Adongo, C. A., Anuga, S. W., & Dayour, F. (2015). Will they tell others to taste? International tourists' experience of Ghanaian cuisines. Tourism Management Perspectives, 15, 57–64. Aguiló, E., Rosselló, J., & Vila, M. (2017). Length of stay and daily tourist expenditure: A joint analysis. Tourism Management Perspectives, 21, 10–17. Ajiferuke, I., & Famoye, F. (2015). Modelling count response variables in informetric studies: Comparison among count, linear, and lognormal regression models. Journal of Informetrics, 9(3), 499–513. Akyeampong, O. A. (2008). Attractions: The bedrock of tourism development in Ghana. , in: Akyeampong, O., & Asiedu, A. B. (Eds.). (2008). Tourism in Ghana: A modern synthesis (1094670516649057). Accra: Assemblies of God Literature, 1–23. Alegre, J., Mateo, S., & Pou, L. (2011). A latent class approach to tourists’ length of stay. Tourism Management, 32(3), 555–563. Alén, E., Nicolau, J. L., Losada, N., & Domínguez, T. (2014). Determinant factors of senior tourists' length of stay. Annals of Tourism Research, 49, 19–32. Arnould, E., & Price, L. (1993). River magic: Extraordinary experience and the extended service encounter. Journal of Consumer Research, 20(1), 24–45. Badu-Baiden, F., Boakye, K. A., & Otoo, F. E. (2016). Backpackers' views on risk in the Cape Coast-Elmina area of Ghana. International Journal of Tourism Sciences, 16(1– 2), 1–14. http://dx.doi.org/10.1080/15980634.2016.1185308. Barros, C. P., & Machado, L. P. (2010). The length of stay in tourism. Annals of Tourism Research, 37, 692–706. Barnes, S. J., Mattson, J., & Sorensen, F. (2016). Remembered experiences and revisit intentions: A longitudinal study of safari park visitors. Tourism Management, 57, 286–294. Boakye, K. A. (2010). Studying tourists' suitability as crime targets. Annals of Tourism Research, 37(3), 727–743. Boakye, K. A. (2012). Tourists' views on safety and vulnerability. A study of some selected C.A. Adongo et al. Journal of Outdoor Recreation and Tourism 18 (2017) 65–74 73 http://refhub.elsevier.com/S2213-0780(17)30010-5/sbref1 http://refhub.elsevier.com/S2213-0780(17)30010-5/sbref1 http://refhub.elsevier.com/S2213-0780(17)30010-5/sbref2 http://refhub.elsevier.com/S2213-0780(17)30010-5/sbref2 http://refhub.elsevier.com/S2213-0780(17)30010-5/sbref2 http://refhub.elsevier.com/S2213-0780(17)30010-5/sbref3 http://refhub.elsevier.com/S2213-0780(17)30010-5/sbref3 http://refhub.elsevier.com/S2213-0780(17)30010-5/sbref3 http://refhub.elsevier.com/S2213-0780(17)30010-5/sbref4 http://refhub.elsevier.com/S2213-0780(17)30010-5/sbref4 http://refhub.elsevier.com/S2213-0780(17)30010-5/sbref5 http://refhub.elsevier.com/S2213-0780(17)30010-5/sbref5 http://refhub.elsevier.com/S2213-0780(17)30010-5/sbref5 http://refhub.elsevier.com/S2213-0780(17)30010-5/sbref6 http://refhub.elsevier.com/S2213-0780(17)30010-5/sbref6 http://refhub.elsevier.com/S2213-0780(17)30010-5/sbref6 http://refhub.elsevier.com/S2213-0780(17)30010-5/sbref7 http://refhub.elsevier.com/S2213-0780(17)30010-5/sbref7 http://refhub.elsevier.com/S2213-0780(17)30010-5/sbref8 http://refhub.elsevier.com/S2213-0780(17)30010-5/sbref8 http://refhub.elsevier.com/S2213-0780(17)30010-5/sbref9 http://refhub.elsevier.com/S2213-0780(17)30010-5/sbref9 http://dx.doi.org/10.1080/15980634.2016.1185308 http://refhub.elsevier.com/S2213-0780(17)30010-5/sbref11 http://refhub.elsevier.com/S2213-0780(17)30010-5/sbref11 http://refhub.elsevier.com/S2213-0780(17)30010-5/sbref12 http://refhub.elsevier.com/S2213-0780(17)30010-5/sbref12 http://refhub.elsevier.com/S2213-0780(17)30010-5/sbref12 http://refhub.elsevier.com/S2213-0780(17)30010-5/sbref13 http://refhub.elsevier.com/S2213-0780(17)30010-5/sbref13 http://refhub.elsevier.com/S2213-0780(17)30010-5/sbref14 towns in Ghana. Tourism Management, 33(2), 327–333. Box-Steffensmeier, J. M., & Jones, B. S. (2004). Event history modelingA guide for social scientists Cambridge: Cambridge University Press. Brida, J. G., Meleddu, M., & Pulina, M. (2013). Factors influencing length of stay of cultural tourists. Tourism Economics, 19(6), 1273–1292. Butler, G., & Hannam, K. (2014). Flashpacking and automobility. Current Issues in Tourism, 17(8), 739–752. Cameron, A. C., & Trivedi, P. K. (2005). Microeconometrics. Methods and applications USA: Cambridge University Press. Campos-Soria, J. A., Inchausti-Sintes, F., & Eugenio-Martin, J. L. (2015). Understanding tourists' economizing strategies during the global economic crisis. Tourism Management, 48, 164–173. Chen, G., Bao, J., & Huang, S. S. (2013). Developing a scale to measure backpackers' personal development. Journal of Travel Research, 1(10), 1–15. Chung, J. Y., Kyle, G. T., Petrick, J. F., & Absher, J. D. (2011). Fairness of prices, user fee policy and willingness to pay among visitors to a national forest. Tourism Management, 32(5), 1038–1046. Cohen, E. (1979). A phenomenology of tourist experiences. Sociology, 13(2), 179–201. Cohen, E. (1972). Towards a sociology of international tourism. Sociological Research, 39, 164–182. Cohen, S. (2011). Lifestyle travellers: Backpacking as a way of life. Annals of Tourism Research, 38(4), 1535–1555. Crouch, G. I. (2011). Destination competitiveness: An analysis of determinant attributes. Journal of Travel Research, 50(27), 27–45. Dann, G. M. (1977). Anomie, ego-enhancement and tourism. Annals of Tourism Research, 4(4), 184–194. Dayour, F. (2013). Motivations of backpackers in the Cape Coast-Elmina Conurbation Ghana. African Journal of Hospitality Tourism and Leisure, 2(3), 1–13. Dayour, F. (2013). Are backpackers a homogeneous group? A study of backpackers' motivations in the Cape Coast-Elmina conurbation, Ghana. European Journal of Tourism, Hospitality and Recreation, 4(3), 69–94. Dayour, F., Adongo, C. A., & Taale, F. (2016). Determinants of backpackers' expenditure. Tourism Management Perspectives, 17, 36–43. Ferrer-Rosell, B., Martínez-García, E., & Coenders, G. (2014). Package and no-frills air carriers as moderators of length of stay. Tourism Management, 42, 114–122. Frew, G., McGeorge, E., Grant, S., & de Wildt, G. (2016). Hepatitis B: A cross-sectional survey of knowledge, attitudes and practices amongst backpackers in Thailand. Travel Medicine and Infectious Disease, 1, 1–6. Fuchs, G., & Reichel, A. (2011). An exploratory inquiry into destination risk perceptions and risk reduction strategies of first time vs. repeat visitors to a highly volatile destination. Tourism Management, 32(2), 266–276. Ghana Tourism Authority (2014). Tourist statistical fact sheet on Ghana Accra: Ghana Tourism Authority. Gokovali, U., Bahar, O., & Kozak, M. (2007). Determinants of length of stay: a practical use of survival analysis. Tourism Management, 28, 736–746. Gurtner, Y. (2016). Returning to paradise: Investigating issues of tourism crisis and disaster recovery on the island of Bali. Journal of Hospitality and Tourism Management, 1–9. http://dx.doi.org/10.1016/j.jhtm.2016.04.007. Hajibaba, H., Gretzel, U., Leisch, F., & Dolnicar, S. (2015). Crisis-resistant tourists. Annals of Tourism Research, 53, 46–60. Hall, C. M. (2010). Crisis events in tourism: Subjects of crisis in tourism. Current Issues in Tourism, 13(5), 401–417. Hannam, K., & Diekmann, A. (2010). From backpacking to flashpacking: Developments in backpacker tourism research. , in: Hannam, K., & Diekmann, A. (Eds.). (2010). Beyond backpacker tourism: Mobilities and experiences (1094670516649057). Bristol: Channel View Publications, 1–7. Harris, L. C. (2012). Ripping off tourists: An empirical evaluation of tourists' perceptions and service worker (mis) behavior. Annals of Tourism Research, 39(2), 1070–1093. Harris, L. C., & Ogbonna, E. (2002). Exploring service sabotage the antecedents, types and consequences of frontline, deviant, antiservice behaviors. Journal of Service Research, 4(3), 163–183. Henderson, J. C. (2007). Tourism crises: Causes, consequences and management. Routledge. Henderson, J. C. (2003). Terrorism and tourism: Managing the consequences of the Bali bombings. Journal of Travel & Tourism Marketing, 15(1), 41–58. Hübner, A., & Gössling, S. (2012). Tourist perceptions of extreme weather events in Martinique. Journal of Destination Marketing & Management, 1(1), 47–55. Hunter-Jones, P., Jeffs, A., & Smith, D. (2008). Backpacking your way into crisis: An exploratory study into perceived risk and tourist behavior amongst young people. Journal of Travel & Tourism Marketing, 23(2/4), 237–248. Holcomb, J., & Pizam, A. (2006). Do incidents of theft at tourist destinations have a negative effect on tourists' decisions to travel to affected destinations? , in: Mansfeld, Y., & Pizam, A. (Eds.). (2006). Tourism, security and safety: From theory to practice (1094670516649057). Amsterdam: Elsevier, 153–168. Jonas, A., Mansfeld, Y., Paz, S., & Potasman, I. (2011). Determinants of health risk perception among low-risk-taking tourists traveling to developing countries. Journal of Travel Research, 50(1), 87–99. Kahneman, D., & Tversky, A. (1979). Prospect theory: An analysis of decision under risk. Econometrica: Journal of the Econometric Society, 263–291. Kim, J. H. (2014). The antecedents of memorable tourism experiences: The development of a scale to measure the destination attributes associated with memorable experiences. Tourism Management, 44, 34–45. Kim, J., Ritchie, J. R. B., & McCormick, B. (2012). Development of a scale to measure memorable tourism experiences. Journal of Travel Research, 51, 12–25. Kirillova, K., Fu, X., Lehto, X., & Cai, L. (2014). What makes a destination beautiful? Dimensions of tourist aesthetic judgment. Tourism Management, 42, 282–293. Larsen, S., Øgaard, T., & Brun, W. (2011). Backpackers and mainstreamers: Realities and myths. Annals of Tourism Research, 38(2), 690–707. Latham, G. P., & Saari, L. M. (1984). Do people do what they say? Further studies on the situational review. Journal of Applied Psychology, 61, 507–516. Law, R. (2006). The perceived impact of risks on travel decisions. International Journal of Tourism Research, 8(4), 289–300. Lee, K. H., Alexander, A. C., & Kim, D. Y. (2014). A study of geographical distance groups on length of visitors’ stay at local food festival destinations. Journal of Vacation Marketing, 20(2), 125–136. Leiper, N. (2010). Backpacker Tourism. Current Issues in Tourism, 13(4), 393–397. http://dx.doi.org/10.1080/13683501003782566. Loker-Murphy, L., & Pearce, P. L. (1995). Young budget travelers: Backpackers in Australia. Annals of Tourism Research, 22(4), 819–843. Luo, X., Brown, G., & Huang, S. S. (2015). Host perceptions of backpackers: Examining the influence of intergroup contact. Tourism Management, 50, 292–305. Martinez-Garcia, E., & Raya, J. M. (2008). Length of stay for low-cost tourism. Tourism Management, 29(6), 1064–1075. McKercher, B., & Tse, S. M. (2012). Is intention to return a valid proxy for actual repeat visitation? Journal of Travel Research, 51(6), 671–686. Menezes, G. A., Moniz, A., & Vieira, J. C. (2008). The determinants of length of stay of tourists in the Azores. Tourism Economics, 14(1), 205–222. Money, R. B., & Crotts, J. C. (2003). The effect of uncertainty avoidance on information search, planning, and purchases of international travel vacations. Tourism Management, 24(2), 191–202. Nicolau, J. L., & Más, F. J. (2009). Simultaneous analysis of whether and how long to go on holidays. The Service Industries Journal, 29(8), 1077–1092. Oh, H., Fiore, A. M., & Jeoung, M. (2007). Measuring experience economy concepts: Tourism applications. Journal of Travel Research, 46(2), 119–132. Otoo, F. E., Agyeiwaah, E., Dayour, F., & Wireko-Gyebi, S. (2016). Volunteer tourists' length of stay in Ghana: Influence of socio-demographic and trip attributes. Tourism Planning & Development, 1–18. Otoo, F. E., & Amuquandoh, F. E. (2014). An investigation into the experiences of international volunteer tourists in Ghana. Anatolia: An International Journal of Tourism and Hospitality Research, 25(3), 431–443. Pallant, J. (2005). SPSS survival manual: A step by step guide to using SPSS for windows (version 12) New South Wales, Australia: Allen Unwin. Pearce, P. L. (1990). Farm tourism in New Zealand: A social situation analysis. Annals of Tourism Research, 17(3), 337–352. Pine, B. J., & Gilmore, J. H. (1999). The experience economy: Work is theatre & every business a stage. Harvard Business Press. Patterson, P. G., Brady, M. K., & McColl-Kennedy, J. R. (2016). Geysers or bubbling hot springs? A cross-cultural examination of customer rage from eastern and western perspectives. Journal of Service Research, 19(3), 244–259 (1094670516649057). Prayag, G. (2012). Senior travelers' motivations and future behavioral intentions: The case of Nice. Journal of Travel & Tourism Marketing, 29(7), 665–681. Prebensen, N. K., Kim, H. L., & Uysal, M. (2015). Cocreation as moderator between the experience value and satisfaction relationship. Journal of Travel Research (0047287515583359). Pröbstl-Haider, U., Haider, W., Wirth, V., & Beardmore, B. (2015). Will climate change increase the attractiveness of summer destinations in the European Alps? A survey of German tourists. Journal of Outdoor Recreation and Tourism, 1–13. Ryan, C. (1991). Tourism and marketing- a symbiotic relationship? Tourism Management, 12(2), 101–111. Riley, P. J. (1988). Road culture of international long-term budget travelers. Annals of Tourism Research, 15(3), 313–328. Saayman, M., Krugell, W. F., & Saayman, A. (2016). Willingness to pay: Who are the cheap talkers? Annals of Tourism Research, 56, 96–111. Santos, G. E. D. O., Ramos, V., & Rey-Maquieira, J. (2015). Length of stay at multiple destinations of tourism trips in Brazil. Journal of Travel Research, 54(6), 788–800. Salmasi, L., Celidoni, M., & Procidano, I. (2012). Length of stay: price and income semi‐ elasticities at different destinations in Italy. International Journal of Tourism Research, 14(6), 515–530. Scholtz, M., Kruger, M., & Saayman, M. (2015). Determinants of visitor length of stay at three coastal national parks in South Africa. Journal of Ecotourism, 14(1), 21–47. Stanko, E. (2000). Victims R′Us: The life history of “fear of crime” and the politicization of violence. , in: Hope, T., & Sparks, R. (Eds.). (2000). Crime, risk and insecurity . London: Routledge, 13–31. Uriely, N., Yonay, Y., & Simchai, D. (2002). Backpacking experiences. A type and form analysis. Annals of Tourism Research, 29(2), 520–538. von Schirnding, Y. (2002). Health and sustainable development: Can we rise to the challenge? The Lancet, 360(9333), 632–637. Thrane, C. (2015). Research note: The determinants of tourists' length of stay: Some further modelling issues. Tourism Economics, 21(5), 1087–1093. Thrane, C. (2012). Analyzing tourists' length of stay at destinations with survival models: A constructive critique based on a case study. Tourism Management, 33(1), 126–132. Thrane, C. (2016). Students' summer tourism: Determinants of length of stay (LOS). Tourism Management, 54, 178–184. Thrane, C., & Farstad, E. (2012). Nationality as a segmentation criterion in tourism research: The case of international tourists' expenditures while on trips in Norway. Tourism Economics, 18(1), 203–217. Todd, C. (2009). Nature, beauty and tourism. Philosophical issues in Tourism, 154–170. Tung, V. W. S., & Ritchie, J. R. B. (2011). Exploring the essence of memorable tourism experiences. Annals of Tourism Research, 3, 1367–1386. C.A. Adongo et al. Journal of Outdoor Recreation and Tourism 18 (2017) 65–74 74 http://refhub.elsevier.com/S2213-0780(17)30010-5/sbref14 http://refhub.elsevier.com/S2213-0780(17)30010-5/sbref15 http://refhub.elsevier.com/S2213-0780(17)30010-5/sbref15 http://refhub.elsevier.com/S2213-0780(17)30010-5/sbref16 http://refhub.elsevier.com/S2213-0780(17)30010-5/sbref16 http://refhub.elsevier.com/S2213-0780(17)30010-5/sbref17 http://refhub.elsevier.com/S2213-0780(17)30010-5/sbref17 http://refhub.elsevier.com/S2213-0780(17)30010-5/sbref18 http://refhub.elsevier.com/S2213-0780(17)30010-5/sbref18 http://refhub.elsevier.com/S2213-0780(17)30010-5/sbref19 http://refhub.elsevier.com/S2213-0780(17)30010-5/sbref19 http://refhub.elsevier.com/S2213-0780(17)30010-5/sbref19 http://refhub.elsevier.com/S2213-0780(17)30010-5/sbref20 http://refhub.elsevier.com/S2213-0780(17)30010-5/sbref20 http://refhub.elsevier.com/S2213-0780(17)30010-5/sbref21 http://refhub.elsevier.com/S2213-0780(17)30010-5/sbref21 http://refhub.elsevier.com/S2213-0780(17)30010-5/sbref21 http://refhub.elsevier.com/S2213-0780(17)30010-5/sbref22 http://refhub.elsevier.com/S2213-0780(17)30010-5/sbref23 http://refhub.elsevier.com/S2213-0780(17)30010-5/sbref23 http://refhub.elsevier.com/S2213-0780(17)30010-5/sbref24 http://refhub.elsevier.com/S2213-0780(17)30010-5/sbref24 http://refhub.elsevier.com/S2213-0780(17)30010-5/sbref25 http://refhub.elsevier.com/S2213-0780(17)30010-5/sbref25 http://refhub.elsevier.com/S2213-0780(17)30010-5/sbref26 http://refhub.elsevier.com/S2213-0780(17)30010-5/sbref26 http://refhub.elsevier.com/S2213-0780(17)30010-5/sbref27 http://refhub.elsevier.com/S2213-0780(17)30010-5/sbref27 http://refhub.elsevier.com/S2213-0780(17)30010-5/sbref28 http://refhub.elsevier.com/S2213-0780(17)30010-5/sbref28 http://refhub.elsevier.com/S2213-0780(17)30010-5/sbref28 http://refhub.elsevier.com/S2213-0780(17)30010-5/sbref29 http://refhub.elsevier.com/S2213-0780(17)30010-5/sbref29 http://refhub.elsevier.com/S2213-0780(17)30010-5/sbref30 http://refhub.elsevier.com/S2213-0780(17)30010-5/sbref30 http://refhub.elsevier.com/S2213-0780(17)30010-5/sbref31 http://refhub.elsevier.com/S2213-0780(17)30010-5/sbref31 http://refhub.elsevier.com/S2213-0780(17)30010-5/sbref31 http://refhub.elsevier.com/S2213-0780(17)30010-5/sbref32 http://refhub.elsevier.com/S2213-0780(17)30010-5/sbref32 http://refhub.elsevier.com/S2213-0780(17)30010-5/sbref32 http://refhub.elsevier.com/S2213-0780(17)30010-5/sbref33 http://refhub.elsevier.com/S2213-0780(17)30010-5/sbref33 http://refhub.elsevier.com/S2213-0780(17)30010-5/sbref34 http://refhub.elsevier.com/S2213-0780(17)30010-5/sbref34 http://dx.doi.org/10.1016/j.jhtm.2016.04.007 http://refhub.elsevier.com/S2213-0780(17)30010-5/sbref36 http://refhub.elsevier.com/S2213-0780(17)30010-5/sbref36 http://refhub.elsevier.com/S2213-0780(17)30010-5/sbref37 http://refhub.elsevier.com/S2213-0780(17)30010-5/sbref37 http://refhub.elsevier.com/S2213-0780(17)30010-5/sbref38 http://refhub.elsevier.com/S2213-0780(17)30010-5/sbref38 http://refhub.elsevier.com/S2213-0780(17)30010-5/sbref38 http://refhub.elsevier.com/S2213-0780(17)30010-5/sbref38 http://refhub.elsevier.com/S2213-0780(17)30010-5/sbref39 http://refhub.elsevier.com/S2213-0780(17)30010-5/sbref39 http://refhub.elsevier.com/S2213-0780(17)30010-5/sbref40 http://refhub.elsevier.com/S2213-0780(17)30010-5/sbref40 http://refhub.elsevier.com/S2213-0780(17)30010-5/sbref40 http://refhub.elsevier.com/S2213-0780(17)30010-5/sbref41 http://refhub.elsevier.com/S2213-0780(17)30010-5/sbref41 http://refhub.elsevier.com/S2213-0780(17)30010-5/sbref42 http://refhub.elsevier.com/S2213-0780(17)30010-5/sbref42 http://refhub.elsevier.com/S2213-0780(17)30010-5/sbref43 http://refhub.elsevier.com/S2213-0780(17)30010-5/sbref43 http://refhub.elsevier.com/S2213-0780(17)30010-5/sbref44 http://refhub.elsevier.com/S2213-0780(17)30010-5/sbref44 http://refhub.elsevier.com/S2213-0780(17)30010-5/sbref44 http://refhub.elsevier.com/S2213-0780(17)30010-5/sbref45 http://refhub.elsevier.com/S2213-0780(17)30010-5/sbref45 http://refhub.elsevier.com/S2213-0780(17)30010-5/sbref45 http://refhub.elsevier.com/S2213-0780(17)30010-5/sbref45 http://refhub.elsevier.com/S2213-0780(17)30010-5/sbref46 http://refhub.elsevier.com/S2213-0780(17)30010-5/sbref46 http://refhub.elsevier.com/S2213-0780(17)30010-5/sbref46 http://refhub.elsevier.com/S2213-0780(17)30010-5/sbref47 http://refhub.elsevier.com/S2213-0780(17)30010-5/sbref47 http://refhub.elsevier.com/S2213-0780(17)30010-5/sbref48 http://refhub.elsevier.com/S2213-0780(17)30010-5/sbref48 http://refhub.elsevier.com/S2213-0780(17)30010-5/sbref48 http://refhub.elsevier.com/S2213-0780(17)30010-5/sbref49 http://refhub.elsevier.com/S2213-0780(17)30010-5/sbref49 http://refhub.elsevier.com/S2213-0780(17)30010-5/sbref50 http://refhub.elsevier.com/S2213-0780(17)30010-5/sbref50 http://refhub.elsevier.com/S2213-0780(17)30010-5/sbref51 http://refhub.elsevier.com/S2213-0780(17)30010-5/sbref51 http://refhub.elsevier.com/S2213-0780(17)30010-5/sbref52 http://refhub.elsevier.com/S2213-0780(17)30010-5/sbref52 http://refhub.elsevier.com/S2213-0780(17)30010-5/sbref53 http://refhub.elsevier.com/S2213-0780(17)30010-5/sbref53 http://refhub.elsevier.com/S2213-0780(17)30010-5/sbref54 http://refhub.elsevier.com/S2213-0780(17)30010-5/sbref54 http://refhub.elsevier.com/S2213-0780(17)30010-5/sbref54 http://dx.doi.org/10.1080/13683501003782566 http://refhub.elsevier.com/S2213-0780(17)30010-5/sbref56 http://refhub.elsevier.com/S2213-0780(17)30010-5/sbref56 http://refhub.elsevier.com/S2213-0780(17)30010-5/sbref57 http://refhub.elsevier.com/S2213-0780(17)30010-5/sbref57 http://refhub.elsevier.com/S2213-0780(17)30010-5/sbref58 http://refhub.elsevier.com/S2213-0780(17)30010-5/sbref58 http://refhub.elsevier.com/S2213-0780(17)30010-5/sbref59 http://refhub.elsevier.com/S2213-0780(17)30010-5/sbref59 http://refhub.elsevier.com/S2213-0780(17)30010-5/sbref60 http://refhub.elsevier.com/S2213-0780(17)30010-5/sbref60 http://refhub.elsevier.com/S2213-0780(17)30010-5/sbref61 http://refhub.elsevier.com/S2213-0780(17)30010-5/sbref61 http://refhub.elsevier.com/S2213-0780(17)30010-5/sbref61 http://refhub.elsevier.com/S2213-0780(17)30010-5/sbref62 http://refhub.elsevier.com/S2213-0780(17)30010-5/sbref62 http://refhub.elsevier.com/S2213-0780(17)30010-5/sbref63 http://refhub.elsevier.com/S2213-0780(17)30010-5/sbref63 http://refhub.elsevier.com/S2213-0780(17)30010-5/sbref64 http://refhub.elsevier.com/S2213-0780(17)30010-5/sbref64 http://refhub.elsevier.com/S2213-0780(17)30010-5/sbref64 http://refhub.elsevier.com/S2213-0780(17)30010-5/sbref65 http://refhub.elsevier.com/S2213-0780(17)30010-5/sbref65 http://refhub.elsevier.com/S2213-0780(17)30010-5/sbref65 http://refhub.elsevier.com/S2213-0780(17)30010-5/sbref66 http://refhub.elsevier.com/S2213-0780(17)30010-5/sbref66 http://refhub.elsevier.com/S2213-0780(17)30010-5/sbref67 http://refhub.elsevier.com/S2213-0780(17)30010-5/sbref67 http://refhub.elsevier.com/S2213-0780(17)30010-5/sbref68 http://refhub.elsevier.com/S2213-0780(17)30010-5/sbref68 http://refhub.elsevier.com/S2213-0780(17)30010-5/sbref69 http://refhub.elsevier.com/S2213-0780(17)30010-5/sbref69 http://refhub.elsevier.com/S2213-0780(17)30010-5/sbref69 http://refhub.elsevier.com/S2213-0780(17)30010-5/sbref70 http://refhub.elsevier.com/S2213-0780(17)30010-5/sbref70 http://refhub.elsevier.com/S2213-0780(17)30010-5/sbref71 http://refhub.elsevier.com/S2213-0780(17)30010-5/sbref71 http://refhub.elsevier.com/S2213-0780(17)30010-5/sbref71 http://refhub.elsevier.com/S2213-0780(17)30010-5/sbref72 http://refhub.elsevier.com/S2213-0780(17)30010-5/sbref72 http://refhub.elsevier.com/S2213-0780(17)30010-5/sbref72 http://refhub.elsevier.com/S2213-0780(17)30010-5/sbref73 http://refhub.elsevier.com/S2213-0780(17)30010-5/sbref73 http://refhub.elsevier.com/S2213-0780(17)30010-5/sbref74 http://refhub.elsevier.com/S2213-0780(17)30010-5/sbref74 http://refhub.elsevier.com/S2213-0780(17)30010-5/sbref75 http://refhub.elsevier.com/S2213-0780(17)30010-5/sbref75 http://refhub.elsevier.com/S2213-0780(17)30010-5/sbref76 http://refhub.elsevier.com/S2213-0780(17)30010-5/sbref76 http://refhub.elsevier.com/S2213-0780(17)30010-5/sbref77 http://refhub.elsevier.com/S2213-0780(17)30010-5/sbref77 http://refhub.elsevier.com/S2213-0780(17)30010-5/sbref77 http://refhub.elsevier.com/S2213-0780(17)30010-5/sbref78 http://refhub.elsevier.com/S2213-0780(17)30010-5/sbref78 http://refhub.elsevier.com/S2213-0780(17)30010-5/sbref79 http://refhub.elsevier.com/S2213-0780(17)30010-5/sbref79 http://refhub.elsevier.com/S2213-0780(17)30010-5/sbref79 http://refhub.elsevier.com/S2213-0780(17)30010-5/sbref80 http://refhub.elsevier.com/S2213-0780(17)30010-5/sbref80 http://refhub.elsevier.com/S2213-0780(17)30010-5/sbref81 http://refhub.elsevier.com/S2213-0780(17)30010-5/sbref81 http://refhub.elsevier.com/S2213-0780(17)30010-5/sbref82 http://refhub.elsevier.com/S2213-0780(17)30010-5/sbref82 http://refhub.elsevier.com/S2213-0780(17)30010-5/sbref83 http://refhub.elsevier.com/S2213-0780(17)30010-5/sbref83 http://refhub.elsevier.com/S2213-0780(17)30010-5/sbref83 http://refhub.elsevier.com/S2213-0780(17)30010-5/sbref84 http://refhub.elsevier.com/S2213-0780(17)30010-5/sbref84 http://refhub.elsevier.com/S2213-0780(17)30010-5/sbref85 http://refhub.elsevier.com/S2213-0780(17)30010-5/sbref85 http://refhub.elsevier.com/S2213-0780(17)30010-5/sbref85 http://refhub.elsevier.com/S2213-0780(17)30010-5/sbref86 http://refhub.elsevier.com/S2213-0780(17)30010-5/sbref87 http://refhub.elsevier.com/S2213-0780(17)30010-5/sbref87 The tourism experience-led length of stay hypothesis Introduction Literature review LOS Holiday experience Materials and methods Sample characteristics Results Independent variables Econometric results Discussion Conclusion and implications Acknowledgements References