Sparks, Ch. 2, "Scientific Methods in Media Effects Research"
CONTENT ANALYSIS--study message content
*descriptive, but in "a systematic and rigorous" way
*logical beginning, but does not allow the researcher to draw conclusions about the effects of the content
*objective-different researchers studying the same content should reach the same conclusion
*cannot tell us about the effects of the message
Example 1: Glantz, Kacirk & McCulloch 2004 study of smoking in films. ppt.
Example 2: Primack et al. 2008 study of substance use in songs. ppt
SURVEY--study people
*descriptive
*ask questions of a
*sample of a population of interest
Example 1: Davis and Mares 1998 study of the effects of viewing talk shows on adolescents. ppt.
Example 2: Ahern et al--2004 study of television images and post-traumatic stress after 9-11. ppt.
EXPERIMENT--study people under controlled conditions
*search for possible causal relationships
^correlation
^time order
^rule out extraneous variables/"third variable explanations"
*manipulate the independent variable to look for change in the dependent variable
^random assignment of subjects to experimental conditions
^identical treatment expect for the manipulation
^control
Example 1: Zillman and Knobloch's 2000 study of the effects of mood on music listening choice
Example 2: Peffley, et al., study of race and crime in TV news stories
Content analyses of local and national news report a disproportionate number of stories about crime. They also report that black criminal suspects are overrepresented in the visuals (but not in the narratives). Peffley et al. proposed two hypotheses to test by an experiment.
HYP 1: Regardless of the race-neutral language used by reporters, the visual image of the race of the suspect is likely to make a crucial difference in the way whites respond to the story.
HYP 2: Whites who hold negative stereotypes of African-Americans are likely to respond more negatively to black than to white suspects.
95 Caucasian subjects from a political science class at a large mid-western university were chosen as subjects.
The subjects were given a pre-test in which a test for racial stereotyping was embedded.
not prone to violence _ _ _ _ _ _ _ prone to violence
hardworking _ _ _ _ _ _ _ lazy
rich _ _ _ _ _ _ _ poor
prefer to be self-supporting _ _ _ _ _ _ _ prefer to be on welfare
intelligent _ _ _ _ _ _ _ unintelligent
Then the subjects were divided into 2 groups; both were shown a news segment about the murder of a prostitute. The story was 1:45 and the visual was 7 seconds.
Group 1 (visual was a black suspect)
Group 2 (visual was a white suspect)
Both groups were given a post-test, which included a test of vocabulary and elicited responses to the crime story.
Results:
*Ss with high negative stereotypes thought the black suspect was--
guilty
deserving of a long sentence
would commit crime in the future
and Ss
had an emotional response of fear
Example 3: Bandura's Study at Stanford University Nursery School
Bandura studied 48 male and 48 female subjects at the nursery school. Their ages ranged from 3 to 5 years. Each child was mildly frustrated and then assigned to one of 4 groups. The 3 experimental groups saw aggression against a Bobo doll; the 4th group was a Control group.
G1: a person in the room acts aggressively towards a Bobo doll
G2: a person on film acts aggressively towards a Bobo doll
G3: a cartoon character on film acts aggressively towards a Bobo doll
G4: control
Then the Ss were observed at play.
Results:
*G1 Ss engaged it 83 acts of aggression; G2 Ss engaged in 92 acts of aggression; G3 Ss engaged in 99 acts of aggression and G4 Ss engaged in 54 acts of aggression.
*Males exhibited more aggression than females.
TERMS
cross sectional survey
longitudinal survey
replication
convergence
epidemiological research