Theories are formed to explain or describe something such as relationships, events or behavior. A theory gives a statement on observations and consists of an interrelated, logical set of ideas and models. A structure that could hold or support a theory in a study is called the theoretical framework. It explains why the problem from the study is present. Thus, the theoretical framework is a theory that plays the role as a starting point for conducting a research. Theoretical framework is very important in a research as it helps the researcher to clearly show the variables used in the study and it can help to form the general framework for data analysis. There are two types of variables which are independent variable and dependent variable.
Sampling is a subgroup or a subset of a population (Sekaran, 2003). There are two types of sampling which are non-probability sampling and probability sampling. Probability sampling is the selection of sample based entirely on chance. The researcher has no control on the selection or non-selection of the sample. Probability sampling is named as random sampling. In contrast, non-probability sampling is the selection of sample which is not depends on chance. The samples are selected by the researcher. Non-probability sampling is also known to be non-random sampling. Another term is population. Population is a group of people that have one or more common characteristics which data can be collected and analyzed.
In order to perform power analysis, Gpower statistical program will be used in this research (Erdfelder, Faul, & Buchner, 1996). This Gpower is used to obtain high-precision for the most common statistical tests in behavioural research such as t-tests, F-tests and Chi-tests. For examples, it could be used for independent samples, correlations, multiple correlation and regression and also contingency tables. The program basically computes the power values, sample sizes and alpha and beta values. This program could also show the relation between any two relevant variables graphically and calculate the effect size measures from basic parameters defining the alternative hypothesis.