The generalization is the extent to which your study is true in different settings. The idea of making sure that your study is well represented makes your study findings generalized. If your sample is not representing the sample fairly, then probably the findings cannot be generalized to the entire population, rather generalized to those who participated.
Quantitative
If your study is quantitative, you need to address the issue of generalization in your study. This is an important question by anyone who evaluates your study. If your sample effectively represents the population, then your study findings have good chance of capturing the truth about your subject matter. This makes your study of value to be looked at. On the other hand, if your study sample did not include a good representation of the population, then your findings would not necessarily be true of the population.
The issue of generalization starts when you plan your methodology. Then, it is discussed in your findings. For example, if you use a random sample, or a quota sample, then you may want to compare distribution of the actual responses that were returned to you with those of the population. This will be discussed in the descriptive analysis of the sample in your findings chapter. If none of the commuters responded to your study, then your study is limited to the non-commuters. If a specific college is underrepresented in your study, then probably the findings may not accurately reflect that college.
Case Study
If your study is a case study, you will have some in-depth findings about your case. Then you may want to show how your case can provide a good sample that represents many other similar organizations. This is a generalization issue.
qualitative study
If your study is a qualitative study, then you assume that your findings may be true for a wider population. The thesis or dissertation is a journey of finding the truth about the world. Therefore, as you find the truth, you have to describe how much of this truth is applicable to the world.