COMM 110: EXAM 1
45 Cards in this Set
| Front | Back |
|---|---|
|
Empirical Research in Comm
|
descriptions of observations are expressed predominantly in numerical terms
|
|
Empirical Goals
|
to describe
to predict
to generalize
to explain
|
|
Treatment of Participants
|
Voluntary participation (informed consent)
Freedom from harm (psychological and physical)
Anonymity
Direct or Indirect Benefits (compensation)
|
|
IRB
|
Institutional Review Board
Oversees the treatment of human subjects
All studies that contribute to "generalizable knowledge" must go through IRB
|
|
Experiment
|
The study of the effects of an independent variable (stimulus) manipulated by a researcher on a dependent variable (outcome), while controlling for all other sources of variance on the outcome variable
|
|
Manipulation check
|
to make sure that the stimuli (ex. independent variable) actually operated in the study
|
|
Experiment External Validity
|
generalizability of the findings
|
|
Experiment Internal Validity
|
Ability to establish a causal relationship
Preconditioned for external validity but doesn't guarantee the external validity
|
|
Pre-experimental designs
|
Low- validity
No randomization
|
|
True/Full Experimental Designs
|
High internal and external validity
Randomization
|
|
Experiment Strengths
|
ability to address questions of causation
ability to isolate and study components of complex processes
|
|
Experiment Weaknesses
|
Triviality
Artificiality
Time-consuming to design
|
|
Survey Strengths
|
standardized, and thus systematic and reliable descriptive data
easy to administer
|
|
Survey Weaknesses
|
Once you start administering a survey you can't change it
Standardized
Relying on self-reports
Social desirability- ppl will lie to make selves seem better
|
|
Sampling
|
Selecting a representative sample that accurately reflects characteristics of the population
|
|
Random Sampling
|
Definition: Equal chance of being selected
Pro: Representative Sample
Con: Complicated, time-consuming, need resources
|
|
Nonrandom Sampling
|
Pro: Useful for private, intimate, or deviant matters
Con: Concern for sampling error
Kinda Con: Requires a larger sample size to reduce sampling errors than random sampling
|
|
Sampling Error
|
The gap between what you have; and what the sample tells you about certain things and what the population tells you about certain things
|
|
Time dimension
|
Cross-sectional: only one point in time
Longitudinal: over a long period of time
|
|
Research Questions
|
an interrogative statement inquiring the relationship between two or more variables
|
|
Hypothesis
|
a declarative statement of the predicted relationship between two or more variables
|
|
Null Hypothesis (H0)
|
-definition: An implicit statement that underlies every hypothesis and indicates that there is no difference between the groups or no relationship between the variables being studied.
-what we observe in our sample results purely from chance, and thus it will not be observed in the pop…
|
|
Research Hypothesis (H1)
|
what we observe in our sample is influenced by some non-random cause, and thus it will also be observed in the population
You cannot prove that H1 is tenable by testing H1. You can only support H1 by rejecting H0
|
|
Variables
|
a characteristic of an entity that varies and to which discrete OR continuous numbers may assigned
|
|
Independent Variables
|
The input variable
Cause of an outcome variable
|
|
Dependent Variable
|
The output variable
The predicted/ outcome variable
|
|
Conceptualization
|
What does the variable mean in the study?
Typically in the Lit Review Section; occasionally in the method section
|
|
Operationalization
|
How is the variable measured?
|
|
Likert Scale
|
Composed of statements for which subjects indicate their agreement typically on 5-point or 7-point scales
|
|
Semantic Differential Scale
|
Bounded by pairs of bipolar adjectives typically on 5-point or 7-point scales
No statements
|
|
Measurement Validity
|
refers to how well researchers measure what they intend to measure
|
|
Measure Reliability
|
measuring something in a consistent and stable matter
|
|
Cronbach's Alpha/ Item Reliability
|
This procedure uses the overall relationship among the answers as the reliability coefficient; this is a survey with scaled, close-ended questions, with multiple items
|
|
Cohen's Kappa/ Scott's Pi/ Intercoder Reliability
|
the most common method for assessing reliability is calculating the percentage of agreement between or among the observations of interdependent coders; how reliable the tool/scheme we can put the info into the categories
|
|
confounding
|
failing to manipulate conditions and to control other souces of variance
|
|
Face validity
|
By looking at the content of the measurement items
|
|
Relationship btwn reliability and validity
|
validity guarantees reliability, but reliability does not guarantee validity
|
|
Why is a published measurement scale better?
|
Using a scale that is already made means that it is proven to be valid and reliable
|
|
Sample
|
A subgroup selected from a population or universe
|
|
population
|
All the people who posses a particular characteristic of interest
|
|
statistical significance
|
refers to the probability that the observed result could have occurred randomly if it has no true underlying effect.
|
|
Nominal
|
categories (gender)
|
|
Ordinal
|
ranked categories (class level)
|
|
Interval
|
ordered categories with equal distance between numbers (ex. Temperature)
**no absolute meaningful zero
|
|
Ratio
|
Same as interval but with a meaningful zero. (ex. age) **can't have negative numbers
|
COMM 110: EXAM 1