A hypothesis, defined by Bryman and Bell (2011) is “an informed speculation, which is set up to be tested, about the possible relationship between two more variables”. It is frequently deduced from the theory and is tested (Ibid). When a hypothesis is suggested, statistics are used to determine the probability of the given hypothesis being true (Good, 2000). The process of determination, is therefore, named hypothesis testing (Good, 2000). Bryman (2008) points out that hypothesis testing is a vital stage of conducting quantitative research. Nonetheless, he stresses that, in comparison to social research, specification on hypothesis and hypothesis testing are more likely to be entailed when carrying out experimental research inquiries. Opposed to his opinion, Crewel (2014) proposes that regardless the form of quantitative research inquiries, the most rigorous form of quantitative research is a result of a test of a theory, of which the hypotheses are derived from. Despite the fact that there are disagreements among scholars in regards to under which circumstances hypothesis testing play its most significant role, there is no doubt that during statistical procedures for addressing research questions, hypothesis testing is an inevitable and crucial stage to go through (Zar, 1984). As in results of the hypotheses tests should eventually lead to rejection, confirmation or reformation the theory or model (Newby, 2010).

The very first step of starting a hypothesis testing, is to set out the null and alternative hypotheses (Good, 2000). In order to explain the process and methods of hypotheses testing, the author would firstly address the differences between these two forms of hypotheses that are used. A null hypothesis stands for a hypothesis which stipulates that there is no relationship between two variable in the population (Bryman and Bell, 2011). In other words, they are irrelevant. For example, if the researcher aims to find out the relationship between fear of failure and the tendency of academic procrastination among students in hotel and tourism management institutes. A null hypothesis would then be: H0: Fear of Failure has no effect on tendency of academic procrastination among students in hotel and tourism management institutes. On the other hand, based on prior literature on the topic which has a tendency for suggesting a potential result, alternative hypotheses construct informed predictions about expected outcomes (Crewel, 2014). Referring to the prior example mentioned, an alternative hypothesis would be H1: Fear of Failure has positive effects on tendency of academic procrastination among students in hotel and tourism management institutes or H2: Fear of Failure has negative effects on tendency of academic procrastination among students in hotel and tourism management institutes. Due to the fact that null hypothesis is a statement of no difference (relationship), is it generally the null hypothesis that is referred to when a hypothesis testing is conduced (Sridharan, 2015). Reasons behind are that if a null hypothesis is tested as invalid, the alternative hypothesis would then be accepted tentatively with a conclusion stating that there is a difference or relationship between the two discussed variables. Only one hypothesis needs to be tested if the chosen is a null hypothesis, vice versa, all alternative hypotheses need to be tested in order to be accepted (Good, 2000). Apparently, it is less demanding proving a null hypothesis wrong than proving all the other alternative hypotheses right. Following the construction of null and alternative hypotheses, establishing the statistical significance level is the second step needs to be carried out. Tests of statistical significance plays influential roles in the process of hypothesis testing. Even thought the word “significance” in the term tends to imply the importance of the results, it does not necessarily indicate that the findings are intrinsically important or substantively significant. Level of statistical significance is solely and directly related to how confident a researcher can be about his or her result deriving from the study in regards to the generalisability of the sample to the population from which they were chosen (Bryman and Bell, 2011). Levels of significance are being considered as probability levels. To be specific, the level of probability of rejecting the null hypothesis that is set up before head, when it is actually being expected to be confirmed (Sridharan, 2015). By testing the level of statistical significance, the researcher would then be able to establish the degree of risk that he or she may reject the null hypothesis. “The p-value is the probability of obtaining at least as extreme results given that the null hypothesis is true whereas the significance level α is the probability of rejecting the null hypothesis given that it is true” (Sandra, 2007). Conventionally, the universal acceptance of the level of statistical significance is at its maximum of p < 0.05, which is therefore, considered as a standard significant level. When p value is less then 0.05, it simply indicates that there are fewer than 5 out of 100 chances the researcher would have a sample that shows a relationship when there is not one in the population (Bryman and Bell, 2011). In other words, when the statistical significance of the findings is either equals to or less than 0.05, the researcher could reject the null hypothesis and accept the alternative ones. Conversely, the researcher would fail to reject the null hypothesis. (ibid). Once researchers have had the level of statistical significance of the findings tested, computing the test statistic would be the following stage. Last but not least, researcher need to make eventual decision and interpreting the results. Researcher establish hypotheses which are derived from theories based on the existing literature and testing of these hypotheses are being conducted in order to answer the research questions. As in results of the hypotheses tests should eventually lead to rejection, confirmation or reformation the theory or model (Newby, 2010), hypotheses testing plays a tremendously important role of conducting a statistical research procedure even though here are disagreements among scholars in regards to under which circumstances hypothesis testing corresponds with the rest of the research process most appropriately (Zar, 1984). The author therefore, has discussed the procedure of conducting hypothesis test and cacenpts that are most relevant to it in favour of readers who intends to carry out a statistical research to have a basic understanding of the method. References: Bryman, A. and Bell, E. (2011) Business Research Methods. 3rd ed. Oxford: Oxford University Press. Creswell,J.W. (2014) Research Design: Qualitative, Quantitative, and Mixed Methods Approaches. Available from: https://books.google.ch/ [Accessed 31 November 2015]. Good,. P. (2000) Permutation Tests: A Practical Guide to Resampling Methods for Testing Hypotheses. 2nd ed. New York: Springer-Verlag. Newly., P. (2010) Research Methods for Education. Pearson. Sridharan,. R. (2015) Statistics for Research Projects: IAP 2015. Available from: http://www.mit.edu/ [Accessed 31 November 2015]. Zar,.F. (1984) Quantitative Methods (GEO 441) Hypothesis Testing. Available from: http://webspace.ship.edu/ [Accessed 31 November 2015].