Sparks Chapter 2

Correlation coefficients

Best way to determine relationships between two variables is the technique of correlation - when researchers want to know whether two variables are related to each other, they often apply a statistical formula to the date and compute a correlation coefficient. - Correlation coefficients are computed between two variables -if increases in one of the variables tends to go along with increases in the other variable, then the two variables would have a positive correlation -when two variables tend to go along with decreases in the other variable, they would have a negative correlation - no correlation coefficient can exceed +1.0 or -1.0 -typically reported like r= .34 (positive correlation) and r=-.34 (negative correlation)

Pros of meta-analysis

The result of a meta-analysis is that a statistical average over the entire set of studies that is taken to be a more reliable indiction of any effect than the results of a single study alone - valuable in summarizing across a large number of studies and arriving at a conclusions that might not be evident from simply reading over the literature

Statistically significant

a statistical result that is unlikely to be due to chance is -once a researcher has said the correlation is statistically significant the first criteria for a causal relationship has been meet

Cohort study

allows researchers to assess changes in a general cohort group but not in the same individuals -age cohort same ages but at different times -would use when you want to see a an effect in a group not individuals

Replication

refers to the fact that the same result can be observed over and over again

Convergence

refers to the fact that the use of different methods still leads to the same general conclusion

Random assignment

theoretically makes the experimental groups equivalent just prior to their exposure to the experimental manipulation.

Control group

to properly asses the impact of a manipulated variable in an experiement this is set up

manipulation of the variable

treated differently than the random assignment is treated

Meta-analysis

- a methodological approach - is a technique that does not involve making observations on a sample of people -this technique relies on studies that have already been completed and uses those studies as the data for a new overall summary -key to meta-analysis is that way in which this summary of existing studies is constructed - summary is based on a very precise examination of the statistical effects reported in each individual study

Cross-sectional survey

- a survey of a single sample of a population - it occurs at a single point in time and involves a single sample -often researchers are not satisfied with a single sample -would use when not looking for changes over time

Cons of a meta-analysis

- some published studies do not contain the statistical details contacted to provide the missing information so the study gets dropped - research community would be in deep trouble if everyone decided to simply do a meta-analyses for the rest of their careers - unless basic surveys and experiments are still conducted and reported, meta-analyst would have nothing to look at

Longitudinal survey

-When researchers have questions that cannot be answered unless a separate questionnaire is administered at more than one point in time - when a survey that projects more than one sample taken as different points in time - there are 3 kinds of longitudinal surveys: trend study, cohort study, panel study,

The experimental method

-experiments in mass comm gather data directly from people -experiments seek to gather these data under controlled conditions -should be able to meet the criteria for causal relationship

trend study

-in a trend study, individuals who are surveyed the first time are not the same individuals who are surveyed the second time.

The sample survey

-in most cases, it is based on a random sample of some larger population of interest -most familiar to average citizen - in comm, surveys are invaluable in helping us to understand people's media habits -through survey research, scholars have a good sense of how much TV the average person watches, what types of content, etc - the survey method is an excellent method for getting descriptive insight into a given phenomenon -good at exploring relationships between different variables

Panel study

-long term panel studies are rare because of the time and effort involved to keep track of the same individuals over time -study is important in telling trends that no other method can detect -short term panel designs are more frequent -researchers design short term panels during political campaigns to track attitudinal changes that might take place in response to media changes and other events

Dependent variable

-once an experimenter has identified a variable (independent) that is thought to be the cause for another variable (dependent variable) the strategy is to manipulate the independent variable to create more than one experimental condition

Criteria for causal relationships

1. correlation must be statistically significant 2. the time order of the two variables( the variable doing the causing must precede the variable that is caused) 3. must establish hat the observed relationship is not due to some unmeasured variable casually related to both of the other - all possible third-varibale explanations must be eliminated

Types of surveys

cross-sectional survey, longitudinal survey