Anderson's Media Research Methods, Ch. 2 - 6

Anderson Ch. 2

In studying the media, various entities can be understood or put into categories based on their properties, processes, consequences, and character. Properties can be studied through basic data collection (surveys, statistics) or immersive engagement (close reading, participant observation). Studying processes requires either sequential measurement methodologies or longer-form interpretive work (protocol analysis, interviews). Consequence is studied in a way that attempts to establish cause and effect, and this can be done through experimentation or participant observation. For character, populations need to be studied, and so surveys, rhetorical criticism, and cultural studies can work to answer questions.

Anderson Ch. 3

Methodology should be public, evidence based, and rule directed, and the studies that result should generate public knowledge. Public knowledge is not necessarily true, but it has passed enough standards not to be deemed false. Methodologies are developed in response to theories; without theory, there is no way to create a systematized method.

It is easier to find problems if you keep current with the discipline’s research. This should be done on a global scale, not limiting it to one’s home country. After doing preliminary research, the problem can be refined to express its domain and components. Methods can come at the problem from deductive (general sample to prove a point/theory) or inductive (single case generalized).

When writing up research, there should be a literature review (shows where in the literature there is material missing) and a problem statement (What is/How does/Why does/What good is?). The literature review should essentially point to the problem statement, showing why it is necessary to answer it.

Anderson Ch. 4

After establishing the problem statement, the researcher must move on to method. Empirical methods use the observation of real practice/culture/material in order to build arguments. It can be categorized as either descriptive, field/life-world, or laboratory.

Metric empiricism is concerned with the measurement and observation of variables. It can be accomplished through observation, survey methods, and experimental methods. In some cases, they can work in tandem with other types of research. Interpretive empiricism is aimed at the study of human action, and it includes reconnaissant observation (short-term), participant observation (long term), critical ethnography, and performance ethnography.

There are certain methods that blur the distinction between metric and interpretive empiricism, including textual analysis, discourse analysis, cultural-critical analysis, and dialogic analysis. These are different from mixed methods, which include multiple methodologies from both metric and interpretive methods.

Anderson Ch. 5

Data must be accompanied with an argument because, otherwise, it is largely useless. Before data can be used, the researcher must implicitly make several assumptions, including what constitutes the media/our text and the nature of the audience/public. We also make assumptions about culture (where and how it is constituted), the way in which individuals relate to society and society to culture, and the role or existence of agency.

Once assumptions are taken into account, a theory must be chosen. The major categories of theory are cognitivism (behavior is dictated from within the individual), social action (behavior is influenced from outside forces), psychoanalytical-semiotic (understanding, meaning, and discursive methods), and critical issue theories (Marxism, feminism, race studies). Each area of theory can be addressed to the psychological, sociological, or cultural level, and furthermore, they can target topics falling into media, content, message, audience, or performance. These theories fall into major epistemic categories as well, including Cartesianism, modernism, and postmodernism.

Depending on whether you are using metric or interpretive methods, the appropriate engagement, evidence, claim, and warrants for trustworthiness are going to differ. In metric work, evidence is characterized by measurable observations of variable manipulation. Trustworthiness is phrased in measurements of reliability, precision, accuracy, and validity.

For interpretive work, evidence is based on the researcher’s coherent and resonant narrative. Their ability to do so will be based on their degree of observation and participation in the studied population. They must be able to link warrants to all of their claims, demonstrating why their evidence is sufficient for the proposition. Trustworthiness for interpretive scholars comes out of coherence, resonance, and vraisemblance (the recognition of the narrative as plausible).

Anderson Ch. 6

Once researchers have research questions, they need to be able to translate their work into hypotheses. For each type of question (What? How? Why? So what?) there are appropriate forms of response. Hypotheses usually appear in metric work, using the theoretical constructs appropriate for the measures and methods of the study. They are not nearly as common in interpretive work.

It is a good idea to create research questions that are anchored both in the real world and the academic, theoretical world. This helps establish warrants for the research. Community-based research is one way to get at both types of problems.

All research work that deals with subjects rather than solely texts should anticipate working with the IRB. The board assesses whether the work is competent scholarship (whether they can really find out what the study sets out to find). They also are concerned with informed consent of participants.