Content analysis


return to the course schedule

ED 714 Qualitative Research Methods in Education
Summer 2002
Copyright - Antonia D'Onofrio - 2000/2001/2002

 
Introduction

 

 

This is a problem focused web page that is intended to give you an overview of the procedures involved in content analysis.

Content analysis is a method for breaking down natural text.  The breakdown or analysis divides text into units of analysis.  It tries to find classes of information,  patterns that group the units together logically according to shared attributes. The patterns usually share meaning - perhaps emotional, connotative, semantic, functional regularities.
Here are some terms that are important to the process of content analysis.

How is a theme different from a pattern?  How is a theme different from a class or a category?


Steps in a nutshell

Practice exercise

Here is an excerpt from an essay written by a computer hacker.  My initial reading of this passage includes the impression that its author experience  a strong sense that hackers were disadvantaged by their lack of organization in the face of deliberate and organized law enforcement.

 
Unlike the efforts of the Chicago Computer Fraud and Abuse Task Force, "Operation Sundevil" was not intended to combat "hacking" in the sense of computer intrusion or sophisticated raids on telco switching stations. Nor did it have anything to do with hacker misdeeds with AT&T's software, or with Southern Bell's proprietary documents.

Instead, "Operation Sundevil" was a crackdown on those traditional scourges of the digital underground: credit card theft and telephone code abuse. The ambitious activities out of Chicago, and the somewhat lesser known but vigorous anti hacker actions of the New York State Police in 1990, were never a part of "Operation Sundevil" per se, which was based in Arizona.

Nevertheless, after the spectacular May 8 raids, the public, misled by police secrecy, hacker panic, and a puzzled national press corps, conflated all aspects of the nationwide crackdown in 1990 under the blanket term "Operation Sundevil." "Sundevil" is still the best known synonym for the crackdown of 1990. But the Arizona organizers of "Sundevil" did not really deserve this reputation -- any more, for instance, than all hackers deserve a reputation as "hackers."

There was some justice in this confused perception, though. For one thing, the confusion was abetted by the Washington office of the Secret Service, who responded to Freedom of Information Act requests on "Operation Sundevil" by referring investigators to the publicly known cases of Knight Lightning and the Atlanta Three. And "Sundevil" was certainly the largest aspect of the Crackdown, the most deliberate and the best organized. As a crackdown on electronic fraud, "Sundevil" lacked the frantic pace of the war on the Legion of Doom; on the contrary, Sundevil's targets were picked out with cool deliberation over an elaborate investigation lasting two full years.

And once again the targets were bulletin board systems.

Boards can be powerful aids to organized fraud. Underground boards carry lively, extensive, detailed, and often quite flagrant "discussions" of lawbreaking techniques and lawbreaking activities. "Discussing" crime in the abstract, or "discussing" the particulars of criminal cases, is not illegal -- but there are stern state and federal laws against cold-bloodedly conspiring in groups in order to commit crimes.

In the eyes of police, people who actively conspire to break the law are not regarded as "clubs," "debating salons," "users' groups," or "free speech advocates." Rather, such people tend to find themselves formally indicted by prosecutors as "gangs," "racketeers," "corrupt organizations" and "organized crime figures."

Note that I have highlighted several words in violet.  My marginal impression is one that picks up impressions of exaggeration or exaggerated law enforcement efforts.  But as I work systematically I will also identify other impressionistic patterns.  The words that correspond to additional patterns will be highlighted in different colors.
 

This is not the only emergent pattern that I am aware of.  There are at least eight others.  Perhaps more.  After I go through the passage and highlight text that seems to match exaggerated law enforcement efforts, I will go through as many more times as there may be patterns or categories, highlighting each set of words or terms in a different color.
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

In this paragraph I am also beginning to pick up a pattern of that I would describe as invisible organization. What units of text probably led me to this inference?


 
When I complete the task of highlighting terms and recording descriptive attributes, I will probably end up with categories that include hacker chaos, exaggeration, aggressive police work, public official organization (rather than underworld organization)  and invisible police organization.  These are temporary and provisional categories.
Most likely I will have many more terms linked to attributes.  I will also have pulled out more patterns than the ones in the chart.  I keep this process up until all the meaningful units have been classified in some way.
My next step will be to connect these categories. Perhaps my theme will be something like:  Deliberation and invisibility - key elements in a successful police effort to overturn disorganized hackers.
Complete the following steps as you analyze the paragraphs in the hacker vignette: