An important first step is to try to tune one’s brain in to seeing things through the prism of statistics. This is not easy though, since human beings are not statistical thinkers. In general, statistical reasoning is based on absolute and relative frequencies and relations between two or more phenomenon. A second step is to try to avoid anxiety and accept the fact that learning statistical methods will take some time and effort. Especially when one is inexperienced, everything feels disconnected and thus incomprehensible. But step by step, things will begin to come together, and some insights and a-ha moments will follow.
In statistics everything is connected. To understand regression analysis one needs to understand the concept of correlation. To understand correlation one needs to know what continuous variables, distributions, and linearity mean. Once the knowledge is built gradually and the relevant steps are taken, it will be much easier to begin to understand more complicated methods, such as regression analysis.
First of all, it is relevant to note that the kind of statistical research done in the social sciences is empirical in its nature, i.e. we are not doing mathematics with numbers, but using the statistical toolbox (i.e. methods) for making observations and inferences about empirical phenomena in a rigorous way. Secondly, all the information that is collected should be in a measurable form, i.e. transferable into numbers. By empirical I did not mean the kind of empiricism without any theoretical concepts that was conducted by the early positivists in the 19th century. Theories have an imperative role in scientific empirical research in formulating research hypotheses and interpreting results.