Learning which type of dissertation you would like to pursue is a part of the first stage when writing a dissertation. Choosing the type of dissertation goes hand in hand with choosing your research topic, because your variables define the methodology you need which consequently defines the type of dissertation you are doing, and this should get clearer as you go along finding the topic you aim to cover because whichever research questions you decide to address, the validity of the answers to those questions depends on the methodology you used.
The most common type of research uses the qualitative paradigm because this type of research offers a rich detail on a topic because its variables are easier to control. These variables refer to the research question, the data needed, the tools needed, and the freedom for educated interpretation. Qualitative studies are further divided into categories, some of which are:
Phenomenology refers to the study where one seeks to present a data on how people perceive and experience a common event. Usually, this method offers the researcher the option to utilize interviews, questionnaires, and observations to draw out accurate findings. For example, you want to find out which teaching approach is best for all types of students in first grade.
Ethnography is somehow similar to phenomenological studies, but ethnographic studies require the researcher to get involved with his respondents or get immersed in the environment under study. Ethnography is often presented in documentaries.
Narrative is the specific version of phenomenology where the researcher follows a single respondent in form of sequential overview related to a specific phenomenon. The researcher can then opt to focus on a specific angle that has immense significance to the research topic.
Grounded theory refers to a study where the researcher attempts to form a theory that explains a phenomenon. Do not confuse grounded theory with phenomenology, because the former aims to identify causes related to the phenomenon whereas the latter aims to present the perspectives of those involved with the phenomenon.
On the other hand, quantitative research depends on figures and statistical results to draw out findings for a specific research topic. Quantitative research is often used to study a larger data-set using standard equations and systematic methods.
Experimental research offers to draw out cause and effect relationships, therefore, allows room for manipulation of the variables. For example, a researcher tries to find out the best ratio of water and fertilizer for a specific amount of soil in growing a certain plant. In this case, the researcher can adjust the volumes of water and fertilizer, while the independent variable soil is locked in a certain amount.
Non-experimental research, on the other hand, does not allow manipulation of variables. This means that the research can only take actual figurative data from the set, therefore, alternative conclusions are present. This is also why most non-experimental studies are correlational studies, and not a concrete cause and effect research. For instance, the researcher aims to find out which gender presents a better class performance. In this case, tests can be implemented, and scores will be quantified using the best formula, and whatever the rule of the chosen formula is, it will be followed (e.g. correlation coefficient is greater or equal to 1 means the relationship is stronger and vice versa).
The key is, again, first identifying the question you would like to address, followed by figuring out the best method to get the answer to that question. Always aim to address the question, not to impress with all the fancy data you can get as you go along your research. Remember, if you want to focus on the nature and quality of your topic, it is most probably qualitative. But, if you wish to study the extent of your topic, it is most probably quantitative. Do not be afraid if you feel like you need to combine methods under both types of research because that is actually possible, and it is called Mixed Research.