New Ways of Studying Media Influence - PowerPoint PPT Presentation

1 / 23
About This Presentation
Title:

New Ways of Studying Media Influence

Description:

Electronic news archives. Social indicator data bases. Approximation of key variables ... Newspaper. Either medium. 2 days before. 54.04. 48.03. 49.43. 1 day ... – PowerPoint PPT presentation

Number of Views:22
Avg rating:3.0/5.0
Slides: 24
Provided by: UNCJ1
Category:

less

Transcript and Presenter's Notes

Title: New Ways of Studying Media Influence


1
New Ways of Studying Media Influence
UNC-CH School of Journalism and Mass
Communication
2
Consumer expectations in freefall
3
The R word in the Times
4
The media recession?
5
Economists r-word index
6
Problems with traditional studies of mass media
influence
  • Individual-level correlations are near zero
  • Duration of observation, at best, is short
  • Causal direction is always questionable
  • Experimental studies are rare and dont solve
    these problems

7
A different approach
  • Move from individual to aggregate or social level
  • Assemble long-term data from existing sources
  • Observe all relevant factors
  • Behavior or effect
  • Media content
  • Reality

8
A different approach
  • Use time-series statistics to sort out direction
    and strength of various factors

9
Problems
  • Finding relevant data
  • But
  • Electronic news archives
  • Social indicator data bases
  • Approximation of key variables
  • High time-series correlations can be misleading

10
Compiling the data set
11
The media recession
12
Method of analysis
  • Simple time-series statistics widely available
  • Sort out direction
  • Control for auto-correlation
  • ARIMA (auto-regressive, integrated moving
    average)
  • RATS (Regression Analysis of Time Series)

13
Interpretation
  • As issue rises, media reporting leads public
    opinion
  • When issue recedes, reporting lags public opinion
  • Media set the agenda but fall behind when opinion
    shifts

14
Campaign 2000
15
Analysis
  • Granger Causality Value of variable over time
    is product of itself at earlier times plus
    external factors
  • Y Yt-1 Yt-2
  • Y Yt-1 Yt-2 Xt-1 Xt-2

16
TV agenda-setting influence on Campaign 2000
17
Interpretation
  • Strong auto-correlation can be misleading
  • Influence of TV on political involvement small
    but can be important

18
Media influence onGulf War Syndrome (GWS)
19
Media coverage and calls to DoD GWS hotline,
1994-1999
20
Calls influence on media (R2)
21
Media influence on calls (R2)
22
Interpretation
  • Both calls and coverage are response to events
  • Difference between print and broadcast partly
    explained by differences in news cycles
  • No evidence of media epidemic

23
Conclusions
  • Evidence of modest media influence
  • But at least as much evidence that media are
    mirrors of events
  • Both media coverage and behavioral response are
    products of other events that drive news and
    public
  • Realistic assessment of media influence
Write a Comment
User Comments (0)
About PowerShow.com