Chapter 8 Flashcards
CA:
Describe phenomena, observe relationships and make predictors
Phenomena empirically to help us do this:
Conceptualisation of inquiry, formulation of research design and data collection plus analysis
Conceptualisation:
- Describe content variables
- Draw inference about content meanings
- Answer RQ, or hypo to about relationships among variables
- Infer from the content to its context of production and consumption
- Answer RQ or hypo about relationship among variables and non-variables
Typology of content studies
- studies only CA
- studies incorporate ca with other methods to explore influences on content
- use ca in conjunction with other methods to explore effect
Content analyis:
- One or more variables across time
multitude of factors observable through CA for their affluence on some content… - CA in conjunction - assess content or as influences on content
- CA and content as independent variable
to lines of research that combined CA and survey methods
CA should not be carried out absent an explicit hypothesis or research question to efficiently guide the design.
-
Lack of explicit
Spurious
Research questions are more tentative
Unable to predict possible outcomes based on theoretical knowledge.
Correlation
Increase or decrease in one variable associated with another
Causation
from correlation is that observed changes may be coincidental
Causal relationship
- Correlation that satisfies the logical conditions for inferring a necessary connection between a change.
condition: time order
hypo cause should precede the effect - Put the third variable under control
- Control of all rival explanations for why changes in two variables are related. Rival: full range of potential and possible alternative explanations in cause effect relation.
Utilitarianism
To answer the investigators question - gold standard for evaluating research design
Good design:
Make explicit and integrate procedures for selecting a sample for analysis, content categories and units in categories, comparison and classes of inference from data
Ideal research design:
Collects a maximum amount of information with minimal expenditure of time and resources.
Evidence is not capable of a dozen other alternative interpretations
Generalisable results
Intolerance for ambiguity
Heart of research design:
Comparison of content that has theoretical importance. Compared across time or among people.
Take advantage of existing data or measurement techniques
= building a body of integrated knowledge in social science research
Conduct a CA:
Conceptualisation: identify the problem, review theory and research, pose specific research questions and hypothesis
Design
Relevant content, formal design, dummy tables, operationalise, specify population and sampling pretest and establish reliability
Analysis: process data, apply statistical procedures, interpret and report results
Conceptualisation
Problem identification and research objective.
Studies purpose:
Pasteur quadrant: study is important for both body of theory and research and for practical problems or social needs.
Incomplete review of existing knowledge:
- Overdepences on web searches
- An exclusion of important journals or all the volumes of those journals 3. An unfamiliarity with scholarship from other fields
- Impatience to get with a project before examining all relevant materials
Formal design of the study:
Actual blueprint for execution of the study. Particular decisions study time frame, data points or comparisons involved
Dummy tables
Hypothetical study outcomes, data collected for study variables and measurement levels > help evaluate.
Heart: content analysis protocol:
Explains how variables are measured and recorded on a coding sheet.
Population of content
Entire set of potential tweets, broadcast programs, documents etc. Representative samples of population. Sometimes not appropriate with samples. Terror attack?
Is quality maximised?
Operational definitions:
Pretested and coders need to be trained. Coder reliability will need testing. Coders should be trained: applying rules and using instructions.
Data collection and analysis
Influence choice of statistical tests:
Level of measurement and type of sample used.
Content variation is affected by five levels of influence.
Five levels: individual media characteristics, media organization routines media organization characteristics, the environment of media organizations, and societal ideology and culture.
Media worker characteristics:
Personnel: direct, creative influence on content. Influencing their work: demographic ones, gender race or other. Sociological research. Political orientation, values and attitudes.
Media organization routines
Routines: repeated patterns of interaction that enable organizations to function. Deadlines for media content production etc. News sources, audienes and processers.
Media organization characteristics
Goals and resources of organizations producing media content, and setting such goals. Rewards and punishments are allocated for achieving such goals. Interactions: with other organizations include resource dependencies and relative power.
Media organization environments
Other organizations and social institutions affecting the work of media organizations
How laws and governmental regulations influence the media organization. Copes with competitors, critics and interest groups?
Societal Idology:
Shoemaker: ideology as a way in which dominant economic interests influence organization environments, characteristics, routines and media workers
Summary:
CA involves conceptualisation, design and execution phases. Research design is the blueprint; the plan specifying how a particular content analysis will be performed to answer a specific research question.
Design: time, comparisons with other media, operationalisation and measurement decisions, sampling, reliability and appropriate statistical analysis. Good research: evaluated in terms of how well it permits answering the research question and fulfilling the study’s purpose.
