Content Analysis

                  Content Analysis
 

Written by
- Kimberly A. Neuendorf in The Content Analysis Guidebook. California: Sage Publications. 2002.
- www.newsimproved.org
- writing.colostate.edu

I. Definition
Content analysis may be briefly defined as the systematic, objective, quantitative analysis of message characteristics. It includes the careful examination of human interactions; the analysis of character portrayals'in TV commercials, films, and novels; the computer-driven investigation of word usage in news releases and political speeches; and so much more.

Content analysis is applicable to many areas of inquiry, with examples ranging from the analysis of naturally occurring language (Markel, 1998) to the study of newspaper coverage of the Greenhouse Effect (Miller, Boone, & Fowler, 1992) and from a description of how the two genders are shown on TV (Greenberg, 1980) to an investigation of the approach strategies used in personal ads (Kolt, 1996). Perhaps, one of the more surprising applications is Johnson's (1987) analysis of Porky Pig's vocalics from a clinical speech therapy standpoint. He examined 37 cartoons, finding that the per-cartoon stuttering ranged from 11.6% to 5 1.4% of words uttered, and certain behaviors were associated with the stuttering (Kimberly A. Neuendorf)

Content analysis is a research tool used to determine the presence of certain words or concepts within texts or sets of texts. Researchers quantify and analyze the presence, meanings and relationships of such words and concepts, then make inferences about the messages within the texts, the writer(s), the audience, and even the culture and time of which these are a part.

Texts can be defined broadly as books, book chapters, essays, interviews, discussions, newspaper headlines and articles, historical documents, speeches, conversations, advertising, theater, informal conversation, or really any occurrence of communicative language. Texts in a single study may also represent a variety of different types of occurrences, such as Palmquist's 1990 study of two composition classes, in which he analyzed student and teacher interviews, writing journals, classroom discussions and lectures, and out-of-class interaction sheets. To conduct a content analysis on any such text, the text is coded, or broken down, into manageable categories on a variety of levels--word, word sense, phrase, sentence, or theme--and then examined using one of content analysis' basic methods: conceptual analysis or relational analysis. (writing.colostate.edu)

II. Reliability
Reliability has been defined as the extent to which a measuring procedure yields the same results on repeated trials (Carmines & Zeller, 1979). When human coders are used in content analysis, this translates to intercoder reliability, or level of agreement among two or more coders. In content analysis, reliability is paramount. Without acceptable levels of reliability, content analysis measures are meaningless. Chapter 7 addresses this important issue in detail. (Kimberly A. Neuendorf)

III. Validity
Validity refers to the extent to which an empirical measure adequately reflects what humans agree on as the real meaning of a concept (Babbie, 1995, p. 127). Generally, it is addressed with the question, "Are we really measuring what we want to measure?" Although,in content analysis, the researcher is the boss, making final decisions on what concepts to measure and how to measure them, there are a number of good guidelines available for improving validity (Carmines & Zeller, 1979). Chapter 6 gives a more detailed discussion. (Kimberly A. Neuendorf)

IV. Generalizability
The generalizability of findings is the extent to which they may be applied to other cases, usually to a larger set that is the defined population from which a study's sample has been drawn. After completing a poll of 300 city residents, the researchers obviously hope to generalize their findings to all residents of the city. Likewise, in a study of 800 personal ads in newspapers, Kolt (1996) generalized his findings to all personal ads in U.S. newspapers in general. He was in a good position to do so because he
(a) randomly selected U.S. daily newspapers,
(b) randomly selected dates for specific issues to analyze, and then
(c) systematically random sampled personal ads in each issue. (Kimberly A. Neuendorf)

V. Replicability
The replication of a study is a safeguard against overgeneralizing the findings of one particular research endeavor. Replication involves repeating a study with different cases or in a different context, checking to see if similar results are obtained each time (Babbie, 1995, p. 21). Whenever possible, research reports should provide enough information about the methods and protocols so that others are free to conduct replications. (Kimberly A. Neuendorf)

VI. Hypothesis Testing
The scientific method is generally considered to be hypothetico-deductive. That is, from theory, one or more hypotheses (conjectural statements or predictions about the relationship among variables) are derived. Each hypothesis is tested deductively: Measurements are made for each of the variables, and relationships among them are examined statistically to see if the predicted relationship holds true. If so, the hypothesis is supported and lends further support to the theory from which it was derived. If not, the hypothesis fails to receive support, and the theory is called into question to some extent. If existing theory is not strong enough to warrant a prediction, a sort of fallback position is to offer one or more research questions.A research question poses a query about possible relationships among variables. In the deductive scientific model, hypotheses and research questions are both posed before data are collected. (Kimberly A. Neuendorf)

VII. Approaches to Content Analysis
The classic Shannon-Weaver (Shannon & Weaver, 1998) model provides the raw framework of source, message, channel, and receiver. Based on this, Berelson (1952) proposed five purposes for content analysis:
(a) to describe substance characteristics of message content (essentially what are described in Chapter 1 as content characteristics),
(b) to.describe form characteristics of message content,
(c) to make inferences to producers of content,
(d) to make inferences to audiences of content, and
(e) to determine the effects of content on the audi-
ence. (Kimberly A. Neuendorf)

VIII. Units
In content analysis, a unit is an identifiable message or message component,
(a) which serves as the basis for identifying the population and drawing a sample,
(b) on which variables are measured, or
(c) which serves as the basis for reporting analyses.
Units can be words, characters, themes, time periods, interactions, or any other result of "breaking up a `communication' into bits". (Kimberly A. Neuendorf)

IX. Content Analysis, Step by Step
1. Define the content you’re going to analyze. Base this definition on your news organization’s editorial goals.
2. Define the audience. Use demographic and other readership data for print. Add reader behavior to these for online.
3. Choose a time frame. What period are you going to measure?
4. Decide what to count. Use specific items, such as story forms, photographs or downloads, that be easily identified.
5. Count. Do the math.
6. Analyze. Apply the numbers to goals and determine success or failure.
7. Repeat regularly. Establish means for ongoing content analysis
(www.newsimproved.org)

X. Steps for Conducting Conceptual Analysis
1. Decide the level of analysis.
2. Decide how many concepts to code for.
3. Decide whether to code for existence or frequency of a concept.
4. Decide on how you will distinguish among concepts.
5. Develop rules for coding your texts.
6. Decide what to do with "irrelevant" information.
7. Code the texts.
8. Analyze your results.
(writing.colostate.edu)