Why content analysis?
What is content analysis?
Systematic and replicable examination of symbols of communication, which have been
assigned numeric values according to valid measurement rules, and the analysis of
relationships involving those values using statistical methods, to describe the
communication, draw inferences about its meaning, or infer from communication to its
context, both of production and consumption
Systematic Generalizable empirical evidence, identification of key
concepts, generation of testable hypotheses. E.g. the scientific
method.
Replicable Definitions are reported carefully so that readers can fully
understand the procedure, evaluate it and its findings, and
repeat it if desired. There must be consistency among each
coder.
Symbols of Different content is interpreted differently based on medium,
communication who is coding, function, etc. This can be avoided with careful
and detailed operationalization.
Numeric values Numeric values are assigned to represent measured
or categories differences. These are not assigned based on counting, but on
some sort of score or rating.
Describing and Description is essential to the early phase of a research
inferring program (e.g. in cases where events are very far from reality,
like hurricane Katrina).
Simple descriptive data invites inference testing (conclusions
about what is/isn’t observed). To draw inferences, the
researcher must be guided by theory
Content is central to processes of communication (figure book centrality of content
model)
Media content connects all kinds of factors of production of content and of facts that
media content might have.
• Cause of (and precondition for) media effects
• Outcome of media production and environment factors (societal factors,
cultural factors, political and so on…)
• Antecedent condition
• Media content
• Media effects
Centrality model of communication content:
, • ● Cause of / prediction for media effects
• ● Outcome of media production and environmental factors
• ● Can be a “consequence” of dominant social ideologies at the time/place
• ● Fundamental to theory building in communication science
Criticisms by Holsti
● Puts too much emphasis on comparative frequency on different symbol’s appearance
• Sometimes specific symbols are crucial and it’s not only about the numbers
● Sometimes problems are chosen only because they are quantifiable
● Hard to distinguish between manifest and latent constructs
• Content analysis makes no claim beyond manifest constructs
Conclusion = content analysis is (accordingly) fundamental to theory building in
Commsci
If you want to make progress in theorizing in commsci, you need also the mediator of
content (p. 35)
Data from content analysis can be linked to data from a survey
• Relevant media content is the independent variable in studies about media
impact
• As a measure of (the opportunity to) exposure, content is superior to
respondent self-report (this last might be biased)
• Measuring content = explicating media impact VS. Not measuring =
speculating
Selection of analysed media content can also serve as (representative) stimulus
material in experiment. But one downside of experiments is low ecological validity,
but if you use content analysis to sample from realistic and real-world content of
media you can actually boost the quality of your experiment
Content analysis offers insight into our mediated reality
Most that we know from the world around us we know from the media
Systematic use of data to understand how media depict/represent reality
• How complete, accurate, realistic is media content? -> media performance
• How does media content originate? Whose reality? -> power relations
Advantages of content analysis over other methods:
• Accessible (as to: data, cost)
• Nonobtrusive, we do not interfere in anyone's life. No need for human
presence
• Nonreactive (-> control), what we study does not really react to the research
process. Mainly for manifest content. With latent contents there is more need
for interpretation, so if our interpretations differ there is some sort of reaction
• Longitudinal in no time. Retrospectively a time machine. We can study
content from ages ago and study the timelines.
• The (relevant) possibilities are endless
Roadmap content analysis
• Formulate research question (hypothesis)
, • Draw a sample (usually)
• Create codebook with coding scheme
• Do coder training (using codebook) (usually: adjust and refine codebook)
• Test reliability of coding
• Repeat step 4 and 5 (usually)
• Code sampled material (collected data)
• Analyse data (answer research question)
• Report study
Research questions
Can be more descriptive or more explanatory in nature
• Descriptive rq used in content analysis -> describe (the variation in) the
presence of characteristics of media content in a given period
• It is still dominant in scientific content analyses. Less the case but still. Very
dominant in practice-oriented content analyses (recurrent: monitoring, trend
research)
• Raise "why questions", so in a sense they are not satisfactory, because they
do not answer those questions
• May lead to answers that stimulate one's curiosity and may make you do
further research
• Raises questions about potential causes (related to media content) and
potential consequences (related to media users)
• Essential as first phase of research program about effects of and on media
content
• Explanatory rq used in content analysis -> explain variation in (a) the
presence of characteristics of media content in a given period of time or (b)
the characteristics of users of media content
• Examples= we study (causal) relationships between content, user, or
production-related characteristics
• Media content can be explained as well as it can be an explanation (see
model figure 1)
Descriptive Explanatory Comparative
Describe the presence of Explain variation in (a) Makes
the descriptive research
characteristics of media content presence of characteristics
questions more explanatory
in a given period in media content within a
given time period or (b)Can compare
Dominant in scientific characteristics of users of • Media categories
studies / practice oriented media content • Content characteristics
content analyses • Time periods
Research about(causal) • Issue contexts
Not ideal for latent constructs
relationships • Environments
Raise “why?” questions butMedia
don’t content can be explained but
answer them can also be an explanation
Stimulate curiosity, can leadE.g.,
to Is the frequency of news about
looking for potential causesvictims
and of a war positively related to
consequences the actual number of war victims?
What is content analysis?
Systematic and replicable examination of symbols of communication, which have been
assigned numeric values according to valid measurement rules, and the analysis of
relationships involving those values using statistical methods, to describe the
communication, draw inferences about its meaning, or infer from communication to its
context, both of production and consumption
Systematic Generalizable empirical evidence, identification of key
concepts, generation of testable hypotheses. E.g. the scientific
method.
Replicable Definitions are reported carefully so that readers can fully
understand the procedure, evaluate it and its findings, and
repeat it if desired. There must be consistency among each
coder.
Symbols of Different content is interpreted differently based on medium,
communication who is coding, function, etc. This can be avoided with careful
and detailed operationalization.
Numeric values Numeric values are assigned to represent measured
or categories differences. These are not assigned based on counting, but on
some sort of score or rating.
Describing and Description is essential to the early phase of a research
inferring program (e.g. in cases where events are very far from reality,
like hurricane Katrina).
Simple descriptive data invites inference testing (conclusions
about what is/isn’t observed). To draw inferences, the
researcher must be guided by theory
Content is central to processes of communication (figure book centrality of content
model)
Media content connects all kinds of factors of production of content and of facts that
media content might have.
• Cause of (and precondition for) media effects
• Outcome of media production and environment factors (societal factors,
cultural factors, political and so on…)
• Antecedent condition
• Media content
• Media effects
Centrality model of communication content:
, • ● Cause of / prediction for media effects
• ● Outcome of media production and environmental factors
• ● Can be a “consequence” of dominant social ideologies at the time/place
• ● Fundamental to theory building in communication science
Criticisms by Holsti
● Puts too much emphasis on comparative frequency on different symbol’s appearance
• Sometimes specific symbols are crucial and it’s not only about the numbers
● Sometimes problems are chosen only because they are quantifiable
● Hard to distinguish between manifest and latent constructs
• Content analysis makes no claim beyond manifest constructs
Conclusion = content analysis is (accordingly) fundamental to theory building in
Commsci
If you want to make progress in theorizing in commsci, you need also the mediator of
content (p. 35)
Data from content analysis can be linked to data from a survey
• Relevant media content is the independent variable in studies about media
impact
• As a measure of (the opportunity to) exposure, content is superior to
respondent self-report (this last might be biased)
• Measuring content = explicating media impact VS. Not measuring =
speculating
Selection of analysed media content can also serve as (representative) stimulus
material in experiment. But one downside of experiments is low ecological validity,
but if you use content analysis to sample from realistic and real-world content of
media you can actually boost the quality of your experiment
Content analysis offers insight into our mediated reality
Most that we know from the world around us we know from the media
Systematic use of data to understand how media depict/represent reality
• How complete, accurate, realistic is media content? -> media performance
• How does media content originate? Whose reality? -> power relations
Advantages of content analysis over other methods:
• Accessible (as to: data, cost)
• Nonobtrusive, we do not interfere in anyone's life. No need for human
presence
• Nonreactive (-> control), what we study does not really react to the research
process. Mainly for manifest content. With latent contents there is more need
for interpretation, so if our interpretations differ there is some sort of reaction
• Longitudinal in no time. Retrospectively a time machine. We can study
content from ages ago and study the timelines.
• The (relevant) possibilities are endless
Roadmap content analysis
• Formulate research question (hypothesis)
, • Draw a sample (usually)
• Create codebook with coding scheme
• Do coder training (using codebook) (usually: adjust and refine codebook)
• Test reliability of coding
• Repeat step 4 and 5 (usually)
• Code sampled material (collected data)
• Analyse data (answer research question)
• Report study
Research questions
Can be more descriptive or more explanatory in nature
• Descriptive rq used in content analysis -> describe (the variation in) the
presence of characteristics of media content in a given period
• It is still dominant in scientific content analyses. Less the case but still. Very
dominant in practice-oriented content analyses (recurrent: monitoring, trend
research)
• Raise "why questions", so in a sense they are not satisfactory, because they
do not answer those questions
• May lead to answers that stimulate one's curiosity and may make you do
further research
• Raises questions about potential causes (related to media content) and
potential consequences (related to media users)
• Essential as first phase of research program about effects of and on media
content
• Explanatory rq used in content analysis -> explain variation in (a) the
presence of characteristics of media content in a given period of time or (b)
the characteristics of users of media content
• Examples= we study (causal) relationships between content, user, or
production-related characteristics
• Media content can be explained as well as it can be an explanation (see
model figure 1)
Descriptive Explanatory Comparative
Describe the presence of Explain variation in (a) Makes
the descriptive research
characteristics of media content presence of characteristics
questions more explanatory
in a given period in media content within a
given time period or (b)Can compare
Dominant in scientific characteristics of users of • Media categories
studies / practice oriented media content • Content characteristics
content analyses • Time periods
Research about(causal) • Issue contexts
Not ideal for latent constructs
relationships • Environments
Raise “why?” questions butMedia
don’t content can be explained but
answer them can also be an explanation
Stimulate curiosity, can leadE.g.,
to Is the frequency of news about
looking for potential causesvictims
and of a war positively related to
consequences the actual number of war victims?
