A research design is the investigator-chosen, overarching study framework that facilitates getting the most accurate answer to a hypothesis or question. Think of research design as similar to the framing of a house during construction. Just as house-framing provides structure and limits to walls, floors, and ceilings, so does a research design provide structure and limits to a host of protocol details.
| Tip. The two major categories of research design are: 1) Non-experimental, observation only and 2) Experimental testing of an intervention. |
DESCRIPTIVE STUDIES
Non-experimental studies that examine one variable at a time.
When little is known and no theory exists on a topic, descriptive research begins to build theory by identifying and defining key, related concepts (variables). Although a descriptive study may explore several variables, only one of those is measured at a time; there is no examination of relationships between variables. Descriptive studies create a picture of what exists by analyzing quantitative or qualitative
data to answer questions like, “What is [variable x]?” or “How often does it occur?” Examples of such one-variable questions are “What are the experiences of first-time fathers?” or “How many falls occur in the emergency room?” (Variables are in italics.) The former question produces qualitative data, and the latter, quantitative.
Descriptive results raise important questions for further study, and findings are rarely generalizable. You can see this especially in a descriptive case study: an in-depth exploration of a single event or phenomena that is limited to a particular time and place. Given case study limitations, opinions differ on whether they even qualify as research.
Descriptive research that arises from constructivist or advocacy assumptions merits particular attention. In these designs, researchers collect in-depth qualitative information about only one variable and then critically reflect on that data in order to uncover emerging themes or theories. Often broad data are collected in a natural setting in which researchers exercise little control over other variables. Sample size is not pre-determined, data collection and analysis are concurrent, and the researcher collects and analyzes data until no new ideas emerge (data saturation). The most basic qualitative descriptive method is perhaps content analysis, sometimes called narrative descriptive analysis, in which researchers uncover themes within informant descriptions. Figure 4 identifies major qualitative traditions beyond content analysis and case studies.
| Alert! All qualitative studies are descriptive, but not all descriptive studies are qualitative. |
Box 1. Descriptive Qualitative Designs
| Design | Focus | Discipline of Origin |
| Ethnography | Uncovers phenomena within a given culture, such as meanings, communications, and mores. | Anthropology |
| Grounded Theory | Identifies a basic social problem and the process that participants use to confront it. | Sociology |
| Phenomenology | Documents the “lived experience” of informants going through a particular event or situation. | Psychology |
| Community participatory action | Seeks positive social change and empowerment of an oppressed community by engaging them in every step of the research process. | Marxist political theory |
| Feminist | Seeks positive social change and empowerment of women as an oppressed group. | Marxist political theory |
asks them to rate 1-4 their knowledge before the class and 1-4 their knowledge now. Then you compare their ratings.
answers to a research question: “Why is the sky blue?”
Quasi-experiments are a lot of work, yet don’t have the same scientific power to show cause and effect, as do randomized controlled trials (RCTs). An RCT would provide better support for any hypothesis that X causes Y. [As a quick review of what quasi-experimental versus RCT studies are, see 
and an experimental group. If the researcher wants to compare health outcomes of smokers with non-smokers, the researcher cannot assign some people to smoke and others not to smoke! Why? Because we already know that smoking has significant harmful effects. (Of course, in a dictatorship, by using the police a researcher could assign them to smoke or not smoke, but I don’t think we wanna go there.)
experimental groups. If the researcher wants to compare health outcomes of
to control & experimental groups. It costs $ & time to get a list of subjects and then assign them to control & experimental groups using random numbers table or drawing names from a hat.
![welcome[1]](https://discoveringyourinnerscientist.files.wordpress.com/2017/03/welcome1.gif)
Herein lies a weak link in the cause-and-effect chain. Quasi- designs are NOT as strong as true experimental designs because something other than our treatment (in this case pet therapy) may have created any difference in outcomes (e.g., anxiety levels). Why? Here’s your answer.
Discovering Your Inner Scientist 