methodology

PRINT

INSTANCEBEGINEDITABLE NAME="DOCTITLE" ASSESSMENT 4 CONTEXT INSTANCEENDEDITABLE

InstanceBeginEditable name="Body"

Methodology

Descriptive Methods

Descriptive methods are one type of non-experimental research. Descriptive methods include surveys that provide a glimpse at the characteristics, thoughts, or attitudes of a sample or interviews (oral questionnaires). The goal of descriptive methods is to provide a snapshot of a particular group—to describe them as they are, right now, at this moment in time.

· Surveys are the most common form of descriptive research. They can be done in a quick, efficient manner with very little interaction between the researcher and the participant. There are many dos and don'ts in survey research; researchers should make sure the questions and the length of the survey are appropriate.

· Interviews can take either an open-ended (unstructured) or closed-ended (structured) approach. Only structured interview can lead to quantitative analysis. Open-ended interviews are much more amenable to qualitative analysis because they contain many non-quantifiable elements. Structured interviews can be rather like an oral administration of a survey.

Correlational Methods

Correlational methods are also descriptive. They try to show a relationship between two variables. The relationship can be either positive (as one variable's value goes up, the other variable's value goes up), or negative (as one variable's value goes up, the other variable's value goes down). These positive and negative descriptions are not indicative of good and bad; they are simply descriptive of a relationship. For example, we desire the negative correlation between exercise and body weight—as exercise goes up, body weight goes down, but we might not desire the negative correlation between shopping and our bank account balance (as shopping goes up, our bank balance goes down). Likewise, a positive correlation can be either a good thing or a bad thing. As piano practice increases, piano skills increase, but as eating ice cream increases, body weight increases.

Correlation does not imply causation—just because something appears to have a correlational relationship does not mean that one variable caused the other variable to change. This is a dangerous assumption that may be better explained by other variables not measured in the study. Correlational research is often the impetus for experimental research where we can test relationships to see if they hold up experimentally.

Correlations are also used to establish reliability. When determining the degree to which we can count on a test to report consistent results, we can use correlations to compare one test administration with another. The degree to which they are correlated informs us of reliability. The Pearson correlation coefficient is the measure of correlation, reported as r.

Qualitative Methods

Qualitative methods are very different from quantitative methods. Qualitative methods use data in the form of interview responses, observations, and a variety of other data in order to create a rich picture of the person, group, or circumstance.

Case Study

Case study is the in-depth study—using multiple methods and data sources—of a single case. Sometimes, a number of cases are studied and reported together. The "case" in a case study is the object of study. One could study a single case (in which a single instance is investigated in depth) or multiple cases (in which a number of instances of the target are studied and then compared with one another).

Program evaluations often are framed as case studies. When a subject is not well described in the scientific literature or is newly emergent, descriptive case studies are often the best way to generate a lot of information about the case on which to base future more tightly focused studies. It is because of the highly descriptive nature of the approach, coupled with the fact that the subject is usually not previously well studied, that case studies seek numerous sources and types of information about the case, including its various contexts.

Case Study Questions

Case study asks questions that involve an intense study of an individual or an organization. The case to be studied is often chosen due to its uniqueness. The case is a single research subject—a person, a program, an event, an activity, or an individual (or individuals). What are the contexts of the case? What are its boundaries? What is involved? What happens? When does it happen? What is like to be in the case?

Ethnography

Ethnography is a descriptive research approach designed for in-depth investigation and description of cultures, cultural groups, large organizations and groupings, and their features. Ethnographers immerse themselves in the culture or organization they are studying, becoming a part of the culture in order to learn about it from the inside out. Consequently, this approach often requires longer timeframes for data collection and ethnographers frequently return a number of times to the sites of their investigations to obtain more data.

An important aspect of this research approach is that the participants (culture) are studied in their natural habitat and social contexts. Individuals are not the unit of analysis for ethnography, although they may be sources of valuable data. Thus, qualitative research questions in social psychology and group psychology often are well answered by ethnographic research.

Types of Questions in an Ethnographic Study

"What are the behavioral patterns of . . . ?" "What is the culture of . . .?" An ethnographic study includes descriptive questions about values, beliefs, and practices of members of the culture under inquiry.

Grounded Theory

Grounded theory is a qualitative research approach that attempts to develop theories of understanding based on data from the real world. The key word is "theory," which in science means an explanatory statement or model based on research evidence. Unlike some other forms of qualitative inquiry, grounded theory attempts to go beyond rich description (which it also strives for) to an explanation of the phenomena of interest.

The second key word is "grounded." This implies that the explanation is derived from the ground, the actual experiences, words, behaviors, and other data obtained from people directly involved or engaged in the topic. For example, if one wished to derive a grounded theory about the effects of childhood abuse on adult functioning, one would gather many kinds of data from persons who had grown up amid child abuse and build the theory of how it affects adult development on the information obtained from those people. Another unique feature of grounded theory is its tendency to return to the ground by taking preliminary insights back to the participants and asking them to further comment on and refine the researcher's conclusions.

The Type of Questions Typically Used to Guide a Study Using Grounded Theory

Process questions about changing experience over time or its stages and phases (for example, "What is the process of becoming . . . ?") or understanding questions (for example, "What are the dimensions of this experience . . . ?"). In grounded theory, the researcher may ask understanding questions, trying to elicit the understanding of the participants about their experiences.

Phenomenology

The key to understanding phenomenology lies in the phrase "lived experience." Put most simply, phenomenology is the study of the lived experience of persons who are going through the phenomenon to be understood. By using the terms "lived experience" and going through, we put the focus squarely on exactly how a phenomenon reveals itself to the experiencing person in all its specificity and concreteness. A phenomenon can be anything that a person experiences—but phenomena (the plural) are defined precisely in their quality of being experienced by someone. A feeling (anger) can be a phenomenon, but a phenomenological study of anger would focus on what it is like to be and to feel anger as an actual, lived experience. Similarly, being hired by a large corporation or being elected to office or losing a loved one in a car accident all are phenomena, but a phenomenological analysis focuses on how the people experiencing them actually experienced them—felt, thought about, perceived, observed, reflected upon them.

The Type of Questions Typically Used to Guide a Study Using Phenomenological Research

"What is the meaning of . . . ?" "What is the experience of . . . ?" "How do people experience and describe . . . ?" "What is the essence of . . . ?" "What is the lived experience of . . .?" "What is it like to experience . . . ?"

Experimental research is, in many ways, the main focus of the research world—it is the most trusted form of research, because when it is done in true experimental form, it allows so much control that one can say with greater confidence than with any other method that the results are true. The goal of experimental research is to isolate a few variables and test their relationship in a controlled environment so that you can eliminate any other possible influences on the behavior. If the relationship holds up under those circumstances (often a laboratory environment), it can be said that one variable causes another. Experimental research is the only type of research where causation can be claimed.

Pre-Experimental Methods

Pre-experimental methods are not the same as true experimental methods and do not carry with them the ability to state causation. Pre-experimental methods set the stage for experimental methods by demonstrating the proposed cause-and-effect situation, though with none of the control of a true experimental study. These are one-shot case studies that have an independent variable and a dependent variable but no control over other factors in order to make certain they had no impact. There is no random selection or random assignment in pre-experimental methods and, thus, conclusions cannot be drawn from this type of research. However, this type of research may lead to true experimental research to test whether the phenomenon is real.

True Experimental Methods

True experimental methods utilize random selection of participants and random assignments of participants to experimental and control groups. These groups allow the researcher to control what happens to each so that he can better understand how the independent variable is impacting the dependent variable. The independent variable is that which differs between the groups, and the dependent variable is that which represents the outcome of how the independent variable impacts the person.

For example, in a study on the effect of music on math skills, music would be the independent variable. Participants would be randomly assigned to an experimental group that listened to music while taking a math skills test or a control group that listened to nothing while taking a math skills test. The dependent variable, or outcome, would be the score on the math skills test. The researcher would tightly control all other circumstances around the testing, from the temperature of the room, to the time of day the test was given, so as to make sure that the only thing that differed was the presence of music. This control would allow the researcher to claim the impact of music on the math skills.

Validity

Experimental research is all about control, and the level of control you exert determines the degree to which the research can be generalized to the population. These two aspects can be measured by looking at how much control was used in the study (internal validity) and the similarity of the research circumstances to the real world (external validity). In a sense these two things are opposites—the more control we exert (the ultimate control being a laboratory environment), the less the circumstances represent the real world. So, in essence it becomes a balance between optimizing internal and external validity.

Quasi-experimental research is somewhat of a compromise. The experimental control required in true experimental designs is not always possible, for a variety of reasons. In some cases, it may be that the groups are preexisting or that it would be unethical to randomly assign people to certain types of groups. If you are studying a group of people who belong to a particular political party, you cannot randomly assign them to one or another party. If you are studying the effects of smoking on mood, you cannot randomly assign people to the smoking condition, as it would be unethical. In these cases, quasi-experimental research is the answer because it does not require the rigor of random assignment to groups, however, with that compromise comes some loss of confidence that the results are in fact due only to the independent variable.

Nonequivalent Control Group Designs

The nonequivalent control group design is the most common quasi-experimental design. It utilizes the same basic set up as an experimental design, with an experimental group and a control group. It also utilizes a pre- and posttest to determine the effect of the independent variables, but there is no random assignment to groups.

If additional restrictions apply to your sample, such as the inability to give a pretest, there are other types of nonequivalent control group designs you can use, such as the static control group comparison or a single-subject design. Each of these brings with it decreasing levels of control, while essentially using the same basic design of studying the impact of one variable on another.

Developmental Designs

Developmental designs are quasi-experimental designs that study a group of people over a long period of time to note changes that occur as a result of development. There are two ways of doing this—the long way and the short way. The long way is referred to as longitudinal research, and the shortcut is cross-sectional research. Longitudinal research may study the behavior of children exposed to particular reading enhancement program to see if that intervention has long term benefits in their high school GPA or college graduation rates. Cross-sectional research may look at several groups of people who are at different developmental stages with some common characteristic, such as a disease or an experience, like being educated in a Montessori school. Looking at people who share that experience but who are now at different periods of time past that experience allows the researcher to make some conclusions about the impact of that experience, without taking 40 years to conduct one study.

References

Kostere, K., & Percy, W. (2006). Qualitative research approaches in psychology. Minneapolis, MN: Capella University.

Salkind, N. J. (2012). Exploring research (8th ed.). Upper Saddle River, NJ: Pearson.