Research Methods Flashcards
Inductive Research/theory
1.) Make an Observation
2.) Choose Methods
3.) Gather Data
4.) Analyze Data
5.) Literature review
6.) Results
-Naturalistic
-There is order at all points-human interaction has patterns
-data to theory/data up/bottom up
-theory is not your friend
Deductive Research/theory
1.) Make an observation
2.) Literature Review
3.) Form Hypothesis
4.) Methods
5.)Gather Data
6.) Analyze data
7.) Write Results
-Formal Procedure
ex: lab experiments
-Predicts Behavior
-Theory to data approach/ top down (theory Down)
-Previous literature is good
-Focuses on quantitative analysis
Pros and Cons of Inductive Research
Advantages
- Go with the flow
-could have many results
-preserves data’s integrity
analyzes behavior includes raw data
Disadvantages
-might not study planned topic
Pros and Cons of deductive research
Advantages
- More control on what is being studied
-structured process
-stats
Disadvantages
- not studying right things
-lab experiments might not be realistic
Major Questions of research methods today?
1.)Media
2.) Technology
3.) Relationships
4.) Identity
What counts as data?
naturalistic data
- video, audio, transcripts
-ethnographic field notes
Contrived data
-data from experiments
-survey data
What do you do with your data?
-Code it or data reduction (Talley marks)
-Stats
-use your data directly
Independent vs dependent variables
Independent Variable will influence change on other variables
Dependent variables are changed by other variables
Research question vs research hypothesis
Hypothesis- will be a statement that ends with a period. Draw Conclusions based on other studies
Research questions- will be phased in question formate that end in a question mark. Not much is know or very broad
Primary vs secondary research reports
Primary Research report are first reporting of studies
Secondary Reports will include second hand account of the primary reports like a text book.
Nominal Measurement Measurement Scales
Classifies a variable into qualitatively different categories like Yes or no.
- They must be mutually exclusive
- Categories must be equivalent
*Categories must be exhaustive
ex- survey Question - Out of the three which is your favorite?
Ordinal Measurement Techniques
*Classifies them using nominal categories
*also ranks order of categories by being greater than or less than
*You do not know the distance between rankings
ex- Race winnings
Likert Measurement Techniques
*Measures extent of a persons beliefs, attitudes or feelings towards something
*include a nutreal or equal point
* equal distance between rankings
ex- Survey question about how you feel about the cold
Semantic Differential Measurement Techniques
- Measures the meaning people associate with concept object or experience
*Uses a scale with two adjective that are opposites at each end (pleasant, Unpleasant)
*Participants rate where they land between the two adjectives
Ex- rank your feelings of public speaking 10 being confident and 1 being not confident
Self-report data
- Ask people to comment on themselves
Advantages
-Efficient
-Allows researchers to study things that cannot be observed
Disadvantages
-what people do and what they believe may be different
-people may not provide accurate information
other-reported data
-Asking other about the participant
Advantages
- may be more accurate then self report data
Disadvantages
-Could be inaccurate
-unmotivated to provide information
-bias
observational data
- When someone observes someone else
advantages
- More accurate
Disadvantages
- Show what people do and not what they believe
-Behaviors observed must match area of interest
Reliability
Consistency and stability
Validity
Accuracy of a study
What is Internal Validity
The accuracy of the conclusion
-How the study was designed
-How study was conducted
-how do your findings support your conclusions
What is External Validity?
Generlizability of findings
- can be applied to other people/texts places or times
Ways to measure reliability
*Multiple admin
- Test retest
-Alternative procedure method
*Single-admin
-Split-half
-Cronbach’s alpha
*Intercoder/interrater/interobserver
Multiple Admin: Test-Retest
*Test retest
Re administer the same measurement producer to the same group of people at different times
Problems
1.) might not be reliable just because there is a change
2.)Time in between re administering will have an impact
3.) Respondents have already been familiarized to test (sensation)
Multiple Admin: Alternative procedure method
*Same people complete another equivalent instrument at the second administration
*Use a coefficient of equivalency to determine the reliability
Problems
1.) Both instruments must be valid and reliable
Single Admin: Split-half
*sperate peoples answers on an instrument into two parts and compare both halves
Problems
1.) people getting tired of answering questions
2.) people getting better at answering questions
Single Admin: Cronbach’s Alpha
*Statistic used to measure realiablity of questions on survery.
*Should be .7 or greater to be relaible
Intercoder/interrater/interobserver
*calculates the percentage of agreement between Intercoder, interrater, interobserver
*allows researchers to be able to pinpoint and change what might be impacting their studies
Threats to internal validity
*How a study is conducted
*Research Participants
*Researchers
Potential Problem with how the study is conducted?
*History
*Sleeper effect
*Sensitization
*Data analysis
What is the history effect?
-Changes in the external environment in a study that causes the participant to have their behavior influenced
What is the sleeper effect?
-Effect that is not immediately apparent
-The results took longer to show up
What is Sensitization?
- Give same test twice and then participants might know what to look for
What threats does DATA ANALYSIS have on internal validity?
-When the incorrect method is used to analysis the data gathered finding might not be Valid
What are potential problems that research participants face
*Hawthorne Effect
*Selection
*statistical regression
*mortality
*maturation
*interparticipant bias
What is the Hawthorne Effect?
The participant know they are being observed
How does SELECTION impact internal validity?
-When you do not have the correct population selected for study. Finding might not be reliable or Valid
What is Statistical Regression?
-Extreme scores go down towards the average (floor and ceiling effect)
What is Morality?
Participants move or die
What is Maturation?
Impacts longer studies where participant become more mature
What is interparticipant bias
-Participants influence each other
Potential Problems Involving Researchers
*Personal attribute effect
*Unintentional expectancy effect
*Observational biases
Personal Attribute effect
-Characteristics like age and race of a researcher influence peoples behavior
Unintentional expectancy effect
- Research unintentionally influence research participants
Observational biases
- Observers knowledge influences observations
- Observers start to become inconsistent with observations
Treats to external validity
*Sampling
*Ecological validity
*Replication
What is Sampling
- a subgroup from a population of people who has a particular characteristics
What are the four types of Random Samples
-Simple
-Systemic
-Stratified
-Cluster
What is a Simple Sample
- each participant is assigned a number then researcher selects from those numbers so there is an equal chance
Problem- Obtaining a list that is represents the population
What is a systemic sample
- You pick a number and every nth person is selected
Problems- Could have a biased sample
What is a stratified Sample
-Categorizes population based on characteristic
Problem- Might be hard to find enough participants to fit each characteristic
What is a Cluster Sample
-Used when a list cannot be obtained
1.) Research selects Custer from population
2.) Research uses cluster from step one and selects individuals from that cluster
Types of Nonrandom Sampling
1.) Convince
2.) Volunteer
3.)Purposive
4.)Quota
5.) Network
What is a Conivence Sample
Asking at random based on availability
Problem- No guarantee that ups will get a sample that is representative
What is a Volunteer Sample?
You have people who self volunteer themselves
Problems- Those who volunteer might be passionate about it
Incentives
What is a Purposive Sample
-Participants are chosen on characteristic
What is a Quota sample
Participants are selected non-randomly on the basis of a characteristic in a population and research sets quota for each characteristic group.
What is a network sample
You ask those who participate for referrals for others
Ecological validity
- Accurately describes what actually occurs in real life circumstances
replication
-Conducting a study that repeats or duplicates in some way a previous study
Exact- Replicated it as closely as possible but will have different participants
Partial- Have some similar things and some things that are different
Conceptual- Has same issues being looked at but different procedure
Ethical issues regarding data
1.) Benefits of Study
2.) Participant consent
3.) confidentiality and Anonymity
4.) Are participants appropriate for the study?
5.) What could be harmful?
6.) consequences
7.) Research Role
8.) Is the design valid?
9.) Will the researcher carryout the study in a valid way?
What is Microanalysis?
Looks at small details that occur within communication by using conversational analysis or Interaction Analysis
Practical Applications of Microanalysis
-Call Centers
- Interviews
-Thearpy
Process of Microanalysis
1.) Pick a interaction/setting
2.) Gather Data
3.) Transcribe data
4.) Analyze data
Advantages and Disadvantages of Microanalysis
Pros-
-Inductive so maximizes possibilities
-Integrity of interactions is preserved
-high ecological validity
-Many studies can be complete from data
-Can look at very tiny details
Cons-
Time consuming to transcribe data
access may be difficult
selection of interaction might be straight forward
What is Ethnography?
-analyzes communication within groups, a communication activity or setting
Process of Ethnography?
1.) Choose something to observe
2.) Negotiate access (involvement and data collection method)
3.)Choose a role
4.) Record data (field Notes)
5.) Conduct Interviews
6.) Analyze data
7.) Interpret results
Pros/ Cons of Ethnographies
Pros-
*Maximizes possibilities
*Preserves integrity of data
*Analyses impact of content on communication
*Finds deeper meaning associated with communication
Cons-
*Hard to Control variables
*Time Commitment
*Could be hard to get access
*Quality can be difficult since it is subjective
*Selection of social action might be biased
*Sampling might be thrown off
Practical Applications of Ethnographies
Study of different cultures
International companies
Corporate mergers
customer satisfaction
What are Surveys?
A way to collect data from a group of people through questions to get information on their beefs feelings and attitudes
Process of Surveys?
1.) Define what you are gathering
2.) Choose population and sampling method
3.) Create questions
4.) Do a pilot study
5.) Distribute Survey
6.) Analyze data
7.) Interpret results
Pro/Cons of Surveys
Pros-
*Allows to see inaccessible behavior
*Pretty straight forward to conduct
*A lot of information in little time
*Gather information from large group
*Can be inexpensive
*deductive so you can study specific variables
Cons-
*Do not deal directly with Communication
*Won’t Accurately report behavior
*social desirability bias
*Can be time consuming and expensive
*Hard to write good questions and format well
*Response rates can be low
*Closed questions may limit findings
*Other areas for bias
What are Experiments?
Control setting to test manipulation of variables to observe effects
Pros/Cons of Experiments
Pros-
*Controlled Environment
*Replicated
*High internal validity
Cons
*Might not accurately reflect really world conditions
*Ethical concerns
*Complex
Process of conducting experiments
1.) Create Hypothesis
2.) Identify variables
3.) Select/create research design
4.) Select a sample
5.) Conduct a pilot study
6.) Administer experiment
7.)Analyze data
8.) Interpret results
Theoretical Saturation in Ethnographies
When you no longer see new patterns in what you are studying
Ethnographic Vs. Survey interviewing
Ethnographic-
to gain deeper understanding of observations
*Informal and Conversational
*Questions can be adaptable to participant
Survey-
To collect standardized information from group of people
*Structured and Formal
*Questions are limited on adaptability
Pros/Cons of Written Questionnaires
Pros
*Efficient
*Standardized
*Can be anonymous
Cons
*Limited Depth
*Low Response Rate
*Research cannot clarify questions
Pros/Cons of Oral Interviews
Pros
-Provide more depth
-Higher response Rate
-See nonverbal
-Questions can be clarified
-Questions can be adaptable to each particpant
Cons
-Time consuming
-Cannot be anonomous
-Expensive
-Research bias
-Can be difficult to standarize
Methods to Increase response rates of surveys
1.) Pre-notification
2.) Personalize it (handwritten)
3.) Incentives
4.) Use green paper
5.) Include a cover letter
Causal Vs. Correlation Relationships
Just because variables correlate does not mean that they are casual
-Casual relationships are where one variable causes another variable to change
-Correlation relationships where changes in variables are linked together.
