METHODS OF COMMUNICATION RESEARCH AND STATISTIC
WEEK 6A: CONTENT ANALYSIS
What is content analysis? A research technique that aims to quantify content in terms of
predetermined categories and in a systematic and replicable manner.
A quantitative, systematic, and objective technique for describing the manifest content of
communications.
A process of systematically sampling, coding and analyzing media content. It can be applied to
both text and visual content.
Step-by-step
1. Develop a RQ and/or H
2. Define the content to be analyzed
3. Sample the content
4. Operationalize variables (define terms)
5. Select units for coding
6. Develop coding scheme
7. Assign each occurrence of a unit in the sample to a code
8. Count occurrences of the coded units
9. Report results, patterns of data and inferences from data
Operationalization
● Variables are the ‘questions’ you ask the content
● Some questions are plain and simple (‘manifest’)
● Latent concepts need multiple indicators (manifest variables) e.g. violence
Sampling
A sample can be generated
● Random (probability sample - simple random, systematic, stratified)
● Select/ Non-random (non-probability sample - convenience or purposive)
Select units for coding
● Unit of analysis: What content do you want to draw conclusions on?
● Sampling unit: What population do you draw your sample from?
● Registration unit: a t what level do you make coding decisions? E.g. every paragraph in
the article or every sentence
● Context unit: how much ‘surrounding information’ do you take into account?
,What could be sensible registration units for video games?
1. Time segments. (10 minutes of playtime) 2. ‘PAT’: Perpetrator engaging in Act against
Target
Unit of analysis? What content do we want to draw conclusions on? 1. 30 CCTV’s coverage of
gymnastics at the 2016 Olympics
Unit of registration? 2. Events (like bars, floor exercise …). The length and gender are being
coded for every event.
Developing and testing a codebook
The code book serves as a guide for coders ( to assign proper scores)
A coding scheme is a form that you use during the coding, a record of each unit
Developing a coding scheme means developing a classification system or categories into which
each sampled unit can be placed. Vehicles, for example, can all be categorized by manufacturer,
color, or state of registration.
The scores can be assigned to dichotomous variables e.g. 0=contains 1= does not contain.
Answer categories must be:
● Clear (provide examples)
● Mutually exclusive (must not overlap)
● Exhaustive (includes all aspects)
, Reliability and validity
Your aim is to develop a code book that leads to consistent coding: no matter who the coder is,
or in what period, the results should be the same each time.
1) One coder: i ntra coder reliability (within) – Same content is coded by the same coder at
two different moments in time
2) Multiple coders: inter coder reliability (between) – Different coders code the same
content and compare afterwards
3) Formal tests: Krippendorff’s Alpha, Lotus
Reliability vs Validity: a tradeoff
Simple codes are often more reliable: ‘What is the article’s general evaluation of the Saudi
government?’ negative OR positive a scale from 1 (very negative) to 7 (very positive)
Chances that coders will assign the same code is higher for the simple code (thus, high
reliability: consistent coding)
But the validity may suffer from simple codes!
Automated content analysis
Automated content analysis is a procedure in that human coders are being supplemented by a
computational method.
Computers are increasingly used in content analysis to automate the labeling (or coding) of
documents. Simple computational techniques can provide descriptive data such as word
frequencies and document lengths.
Pros Cons
● No restrictions on ● Complex tasks are challenging (when it comes to
sample size understanding the meaning of a text, humans are still
● Cheaper in the long superior - e.g. sarcasm)
run ● Generating the algorithm is a subjective human
● Faster process
WEEK 6A: CONTENT ANALYSIS
What is content analysis? A research technique that aims to quantify content in terms of
predetermined categories and in a systematic and replicable manner.
A quantitative, systematic, and objective technique for describing the manifest content of
communications.
A process of systematically sampling, coding and analyzing media content. It can be applied to
both text and visual content.
Step-by-step
1. Develop a RQ and/or H
2. Define the content to be analyzed
3. Sample the content
4. Operationalize variables (define terms)
5. Select units for coding
6. Develop coding scheme
7. Assign each occurrence of a unit in the sample to a code
8. Count occurrences of the coded units
9. Report results, patterns of data and inferences from data
Operationalization
● Variables are the ‘questions’ you ask the content
● Some questions are plain and simple (‘manifest’)
● Latent concepts need multiple indicators (manifest variables) e.g. violence
Sampling
A sample can be generated
● Random (probability sample - simple random, systematic, stratified)
● Select/ Non-random (non-probability sample - convenience or purposive)
Select units for coding
● Unit of analysis: What content do you want to draw conclusions on?
● Sampling unit: What population do you draw your sample from?
● Registration unit: a t what level do you make coding decisions? E.g. every paragraph in
the article or every sentence
● Context unit: how much ‘surrounding information’ do you take into account?
,What could be sensible registration units for video games?
1. Time segments. (10 minutes of playtime) 2. ‘PAT’: Perpetrator engaging in Act against
Target
Unit of analysis? What content do we want to draw conclusions on? 1. 30 CCTV’s coverage of
gymnastics at the 2016 Olympics
Unit of registration? 2. Events (like bars, floor exercise …). The length and gender are being
coded for every event.
Developing and testing a codebook
The code book serves as a guide for coders ( to assign proper scores)
A coding scheme is a form that you use during the coding, a record of each unit
Developing a coding scheme means developing a classification system or categories into which
each sampled unit can be placed. Vehicles, for example, can all be categorized by manufacturer,
color, or state of registration.
The scores can be assigned to dichotomous variables e.g. 0=contains 1= does not contain.
Answer categories must be:
● Clear (provide examples)
● Mutually exclusive (must not overlap)
● Exhaustive (includes all aspects)
, Reliability and validity
Your aim is to develop a code book that leads to consistent coding: no matter who the coder is,
or in what period, the results should be the same each time.
1) One coder: i ntra coder reliability (within) – Same content is coded by the same coder at
two different moments in time
2) Multiple coders: inter coder reliability (between) – Different coders code the same
content and compare afterwards
3) Formal tests: Krippendorff’s Alpha, Lotus
Reliability vs Validity: a tradeoff
Simple codes are often more reliable: ‘What is the article’s general evaluation of the Saudi
government?’ negative OR positive a scale from 1 (very negative) to 7 (very positive)
Chances that coders will assign the same code is higher for the simple code (thus, high
reliability: consistent coding)
But the validity may suffer from simple codes!
Automated content analysis
Automated content analysis is a procedure in that human coders are being supplemented by a
computational method.
Computers are increasingly used in content analysis to automate the labeling (or coding) of
documents. Simple computational techniques can provide descriptive data such as word
frequencies and document lengths.
Pros Cons
● No restrictions on ● Complex tasks are challenging (when it comes to
sample size understanding the meaning of a text, humans are still
● Cheaper in the long superior - e.g. sarcasm)
run ● Generating the algorithm is a subjective human
● Faster process
