Content analysis Holsti Content analysis is any technique

Content analysis (Holsti) Content analysis is any technique for making inferences by systematically and objectively identifying specified characteristics of messages. 1/12/2022 1

Uses of Content Analysis • Describe Communications Content • Tests Hypotheses of Message Characteristics • Compare Media Content to ‘Real World’ • Assess Images of Groups in Society

content analysis • which has sometimes been defined as the drawing of inferences on the basis of appearance or non appearance of attributes in messages. Used a lot in applied social science. i. e. predicting enemy moves according to past action. 1/12/2022 3

Quantitative content analysis • gives strict enumeration of the content examined. . Very rigorous. . . Used perhaps to examine news values among gatekeepers. 1/12/2022 4

Content Analysis • Various types of content design – Characteristics of communication 1/12/2022 5

Content analysis • To make inferences as to the antecedents of communication (encoding) 1/12/2022 6

Content analysis • To make inferences as to the effects communication (decoding) 1/12/2022 of 7

Steps in Conducting CA … üSelect Topic üFormulate Research ü Establish Quantification System … unit of Questions or Hypotheses üDefine the Population that’s being studied ü Train Coders & üSelect an ü Code content Appropriate Sample from Population üConstruct Categories of analysis Conduct a Pilot Study … reliabile and intercoder reliability according to established definitions ü Analyze Data

How do you do a content analysis? • As with any study you must identify what you wish to measure and the purpose of the study. You must determine whether you want to describe the attributes of a message, make inferences about the content by asking who or why and by making inferences about the effects of messages. Once you have identified what you want to measure, you may then proceed with preparing to conduct the content analysis. 1/12/2022 9

Content Analysis • To do a content analysis • Content analysis stands or falls by its categories. – Ask how is the research problem defined in terms of categories? ie. subject matter. . direction. . . origin. . target. . . etc – Definitions of categories requires that they actually represent the elements of the investigators theory. . . : 1/12/2022 10

Content Analysis • To do a content analysis • Content analysis stands or falls by its categories. – – that they be exhaustive, to ensure that every item relevant to the study can be classified: – and that they be mutually exclusive so that no item can be scored more than once within a category set. – 1/12/2022 11

Content Analysis • To do a content analysis • Content analysis stands or falls by its categories. – To insure results which are replicable, the investigator must specify explicitly the indicators that determine which units fall into each category. – Define categories by characterizing their major properties. 1/12/2022 12

Content Analysis • Procedures for content analysis – 1. Select a topic – 2. Decide sample – 3. Define concepts or unites to be counted – 4. Construct categories – 5. Create coding form – 6. Train coders – 7. Collect data 1/12/2022 13

Content Analysis • Procedures for content analysis – 8. Measure intercoder reliability – 9. Analyze data – 10. Report results 1/12/2022 14

What unit of content is to be classified? – Units of Analysis • TV show, movies, websites, press releases – What is your context unit? • Type of newspaper, type of genre, network – What are your recording units? • Frames in a film, five minute increments of a tv show, • Syntactic units…sentences, subjects, verb, adjectives 1/12/2022 15

Enumeration • Now you must deal with the system of enumeration. . The recording unit and the unit of enumeration may be the same. . . quantification of identified unites. 1. time/space. . ie column inches. . . 2. appearance. . . occurrence or nonoccurrence of an attribute 3. frequency. . how many times does unit appear 4. intensity. . how strong or weak is the occurrence. ie. uses of words such as declare, said, urged. . etc 1/12/2022 16

Limitations … ü EX POST FACTO RESEARCH § Can’t show cause & § effect Requires Coding Instructions ü DEALS WITH NUMBERS COUNTING & DEFINITIONS § Categorization is Important § To compare one study with another one must use same categorizations ü SAMPLING IMPORTANT § Time Periods Often § § § Require Multi-Stage Sampling Look for Systemizations in the Data Direct Observation We Do Not Know Why Something Was Done or Wasn’t Done … See If Procedure Works … What is the BASIC CODING UNIT that is being used … that Helps Define the Sample