Research Methodology

According to its aims, the research was defined as descriptive since it describes the features of a certain population as well as establishes relationships between variables through a standardized data collection technique. Its approach is quantitative, once the research used quantification not only in the data collection process, but also in their treatment through statistical methods. The technical procedures used to gather the data enable us to identify this research as a fieldwork - for empirically investigating the place where the event took place and for questioning people whose behavior it intended to know. The field research was applied to an entire universe of medium and big hotels. The data collection tool was a questionnaire with closed questions, organized in two parts. The first one had profile questions for manager/respondent, hotel and Internet use. The second part had perception questions using a non-metric scale for measuring the importance from one to five (on a five-point Likert-type scale), which tried to measure different levels of individual manager importance on the following aspects of Internet use: Efficiency, Access to Internet, Means of Communication, Internet Benefits and Internet Difficulties, containing 44 perception evaluation items. This questionnaire was adapted from Van Hoof and Combrink (1998) and validated for Brazilian reality enabling to confirm the variables that belong to the questionnaire's blocks.

The field research was developed in the city of Natal, capital of Rio Grande do Norte estate in Brazil. The organizational unity chosen was hotel enterprise with more than 40 rooms (for in the researched area they are considered as medium to big-sized) rated on the "Guia Quatro Rodas Brazil." The subjects of the research were the managers who were more involved with Internet use in his hotel.

The "Guia Quatro Rodas Brazil" was used as a reference for rating the categorized researched hotels for two reasons: first for that is the guide which has been currently used by tourists, tour agencies and tour operators for over 36 years in Brazil; secondly because EMBRATUR

(Brazilian Tourism Institute) official organ from the Ministry of Sport and Tourism is still evaluating the hotels according to its new Brazilian Rating System for Hosting Means; once its old system had low credibility for rating (Duarte, 1996), it was not possible to use that new system.

Small properties are out of the study for these sort of facility have less awareness on the importance of adopting technological solutions and also have less financial conditions to invest in IT (Cooper et al., 1998). It is also believed that could interfere in the perception of managers on the impacts of the Internet.

Nine out of the 44 hotels selected were not given the questionnaire application, for these reasons: three of them had already taken the pre-test; four were not rated in the "Guia Quatro Rodas Brazil," one is not working and another one where it was not possible to contact its manager to answer the questions, though it was visited three times. Thirty-five hotels were studied and the managers spent 25 minutes to answer the questionnaire.

This study is considered of unique transversal type, because it extracted only one sample from the interviewees of the target-population and the information was collected only once. The first phase of collecting data refers to pre-test application period developed on the first week of August 2001, in three hotels. After that phase, some due changes were made and the questionnaires were then applied from August 14 to September 13, 2001.

With respect to statistical treatment of data, Microsoft's STATISTICA software for Windows was used to have the descriptive analysis of data, and for inferential results, analysis of variance (ANOVA) and HSD Tukey test were carried out.

In the descriptive analysis, software calculated the frequency table in each one of the variables used in the study and distributed in five parts of the questionnaire. On the first and second parts (Interview/hotel profile and Internet use) it was calculated only the percentages of each of the variables involved. On the remaining parts (Manager Perceptions, Benefits and Difficulties), the variables percentages and averages were calculated.

The analysis of variance was used to check if there was any significant difference on the opinions of hotel managers according to their individual profile (age and experience) and the organizational profile (rate and number of rooms). The ANOVA allows the researcher to compare averages of several different samples (three or more) with data from the ordinal level to determine if population averages from these samples have any significant differences (Montgomery, 1996).

It was also used Tukey's HSD test ("honestly significant difference") to locate where significant differences were located between the averages. Turkey's HSD test is specific for different sized samples, which is the case of this research. Tukey's HSD test is one of the most useful tests for multiple comparisons.