Chapter 10 Content Analysis Content analysis is used

Chapter 10 Content Analysis • Content analysis is used to describe the content of artifacts created by human beings such as texts and audio/visual content © 2017 Taylor & Francis

Content Analysis As an empirical methodology, must adhere to the three basic principles of the scientific method. It must be: 1. Objective 2. Systematic 3. The results must be generalizable © 2017 Taylor & Francis

Content Analysis • This method analyzes the permanent record of a phenomenon, such as: • Texts (e. g. , newspaper articles, speeches, love letters) • Audio/visual content (e. g. , television shows, recorded interpersonal interactions, songs) • We describe by observing and counting selected characteristics of the content under study (the sources of “data” are referred to as texts) © 2017 Taylor & Francis

Conceptualization Identify and define which text will be analyzed: 1. Which data are analyzed? 2. How are they defined? 3. What is the population from which they are drawn? © 2017 Taylor & Francis

The Unit of Analysis • The thing that is actually observed or counted • Selecting the unit of analysis is extremely important because it is used to identify what we will be observing and coding in our study © 2017 Taylor & Francis

Units • In content analysis, part of the conceptualization process is selecting the appropriate text for the study • Unit of observation: the units through which the data are collected © 2017 Taylor & Francis

Units • Sampling unit: the whole independent message used for the basis of sampling • Separable by physically identifiable boundaries • Completely independent of each other • Recording unit: components of the message that can be divided into separate components or categories © 2017 Taylor & Francis

Sampling • In content analysis, sampling is influenced by the type of text we have selected to observe • Target universe: the group of things we are interested in studying (similar to the population in survey and experimental research) • Sampling is a process used to create a sub-group of the target population or universe © 2017 Taylor & Francis

Coding Scheme • Coding scheme: the classification system used to classify messages; provides instructions on how to code or observe the text • Coding: the process by which researchers convert unobserved texts into a form that allows for comparison by statistical analysis © 2017 Taylor & Francis

Coding Scheme Variables • Variables must be: • Mutually exclusive: the variables must have responses that correspond to the data and allow the researcher to categorize the data into one and only one category • Relatively exhaustive: the categories must be inclusive enough for the researcher to be able to record (nearly) all observations © 2017 Taylor & Francis

Two Types of Variables • Manifest variables: can be easily observed and tend to be more objective and reliable • Latent variables: are more subjective and are more challenging to code but can be more valid (represent the concept more accurately) than manifest variables © 2017 Taylor & Francis

Intercoder Reliability • The extent to which independent coders ( the people who observe and categorize the data) evaluate a a characteristic of a message or artifact and reach the same conclusion as the other independent coders © 2017 Taylor & Francis

Computer-aided Coding • There are numerous computer-aided content analysis programs • The most common type is Computer Aided Text Analysis (CATA) • Pros: automated searches; process data quickly and more cheaply than human coders; removes subjective judgments • Cons: not as useful for coding complex concepts or visual samples; failure to recognize multiple meanings and context can contribute to measurement error © 2017 Taylor & Francis