Class Notes (1,100,000)
CA (620,000)
UTSC (30,000)
Sociology (2,000)
SOCA01H3 (600)
Lecture

Lecture 4


Department
Sociology
Course Code
SOCA01H3
Professor
Sheldon Ungar

This preview shows page 1. to view the full 5 pages of the document.
Lecture 4 – Applying methods to violence and the media. Philip’s
studies.
Methods applied to violence and the media
While the question of whether the media causes increased violence seems easy to answer, it is not!
Science:
-Conjectures-Testing -Refutations
Science works by refuting weaker theories rather than proving a specific theory
e.g. “all doves are purple”, you just need to find one dove that is not purple to refute the theory
Can Science predict to the real world?
Science and economics don’t predict well to the real world. Only one person predicted the collapse of the
Soviet Union (everyone else thought they were fine). Weather forecast is wrong many times.
Predictions are more accurate under lab conditions. It is harder predicting about the real world.
Methodology: way of assessing validity and confidence in our knowledge
You start with a question:
e.g. Does violence in the media causes increased violent behaviour
1. Review literature
2. Select a theoretical approach - Select method
Social learning theory
People imitate behaviour of models
-use similar other
-use successful models
Hypothesis
It specifies a relationship between two or more variables. It is a testable statement.
e.g. Observed aggressive behavior on tv leads to increased violence in children
This is more specific than our original question
Research design
Systematic plan to test hypotheses & eliminate alternative explanations
- each design has strengths & weaknesses: hence triangulate designs
Well look at 4 designs today:
Correlational
Natural experiment
Laboratory experiment
Survey
Correlational design
You need to show that as:
Violence on TV increases kid’s violence increases as well, and that the two are related
a-Measure Violence on TV:
Longitudinal content analysis of TV
Random sample of TV shows
Systematic counts of aggression
www.notesolution.com
You're Reading a Preview

Unlock to view full version

Only page 1 are available for preview. Some parts have been intentionally blurred.

Issues:
-Year sample
-What shows you look at adult vs. kids
-How you measure aggression on TV physical vs. verbal
-Difference on definition of aggression 1950s vs. now
b- Show increase in real world kids violence
-longitudinal counts of crime rates
Issues:
-official statistics vs. self reported
-how accurate are official statistics?
-errors in recording
some crime is not reported (petty thefts, car broken into -because of insurance) and this creates errors in
reporting. These errors matter only if error rate changes over time i.e. not constant
e.g. use of PCs changed ways of reporting crime
-changes in definition
Rape vs. sexual assault and harassment, the latter one easier to prove and to get a conviction from
c-Now are the two related over time?
TV violence causes real world kids violence
Vs. Alternative explanations
Is there a third variable affecting TV violence and kids violence?
Alternative explanations:
-demography, number of young men
-level of parental supervision, social control
-availability of weapons (30s-40s Montreals English vs. French fights were fist fights todays fights are
fought with knives and guns)
-changing values
Conclusion: crime rose first, TV violence followed
Natural Experiment design
Randomly assign subjects to:
-Experimental group: high exposure to TV violence
-Control group: low exposure to TV violence
Then compare results
Issues:
-can you control exposure to TV
-number of subjects required need lots of people
-lag timehow long do you have to wait between watching TV and violence appearing. This effect
builds over time thus you cant use a natural experiment
-contaminated by measurement effects
-how do you measure aggression behaviour
e.g. 1 leave $5 on the street for a stranger to pick up are they nicer to people around them after.
e.g. 2 how long does it take for the person behind you to honk during a green light, people in bad cars are
more likely to hit you or wait in the light in front of you longer
Laboratory Experiment design
In a lab, randomly assign subjects to:
-Experimental group: watch a violent film
www.notesolution.com
You're Reading a Preview

Unlock to view full version