Please enable JavaScript.
Coggle requires JavaScript to display documents.
(Communication Research), ] - Coggle Diagram
Communication Research
Research Process
Purpose of Research
Explore / Describe
Used when a topic is not well understood
Asks what exists, what happens, or how people experience something
Explain / Understand
Looks for reasons, meanings, relationships, or processes
Predict
Estimates what is likely to happen based on known patterns or variables
Identifies why the research is being conducted
Research Problem
Why does this study need to happen?
Identifies an issue or gap that needs investigation
Connects to Literature Review
Leads to Purpose Statement
Literature Review
What do we already know?
Not simply a summary of articles
Identifies what we already know
Identifies what we do not know
Identifies disagreements in existing research
Identifies relevant theories and concepts
Shows where our study fits
Research as a Conversation
Listen to what other researchers have said
Identify patterns and disagreements
Find gaps or unanswered questions
Determine where your research fits
Add something meaningful to the conversation
Purpose Statement
What will this study do?
States what the study intends to accomplish
Connects the research problem to the study
Identifies what is being studied
Identifies who or what is being studied
Identifies the context
Identifies the research approach
Quantitative
Examine relationships between X and Y
Compare groups
Test whether X predicts Y
Qualitative
Explore experiences
Understand meanings
Examine how people make sense of something
Mixed Methods
Combines quantitative and qualitative approaches
Different approaches address different parts of the problem
Research Question / Research Hypothesis
What will this study answer?
Research Question
What do I want to know?
May be exploratory and open-ended
Common in qualitative research when the answer is not predicted in advance
Research Hypothesis
What do I expect to find?
Testable prediction about variables or group differences
Common in deductive and quantitative research
Null Hypothesis
Predicts no relationship or no difference
Research / Alternative Hypothesis
Predicts that a relationship or difference exists
Unit of Analysis
Who or what is being studied?
The entity about which conclusions will be made
Examples: individuals, couples, groups, organizations, messages, or events
Do not confuse with the source providing the data
Conceptualization
Refines and specifies abstract concepts
Clarifies what the researcher means by a term
Conceptions
Individual mental images or understandings
Concept
A shared construct or family of conceptions
Construct
Abstract idea that cannot be observed directly, such as trust or anxiety
Conceptual Funnel
Broad interest becomes increasingly focused
Nominal Definition
Working meaning assigned to a concept for the inquiry
Indicators and Dimensions
Concept -> Dimensions -> Indicators
Dimensions are specific facets of a concept
Indicators are observable signs representing a concept
Multiple dimensions reveal conceptual complexity
Interchangeability of Indicators
Reasonable indicators of the same concept should behave consistently
Operationalization
Develops procedures that produce empirical observations
Conceptualization -> Nominal Definition -> Operational Definition -> Measurement
Operational Definition
Precisely states how the concept will be measured
Operationalization Choices
Range of variation
Amount of difference the measure is able to capture
Too little variation can hide meaningful differences
Level of measurement
Determines which mathematical and statistical procedures are appropriate
Single versus multiple indicators
Single indicator is simpler but may miss part of a complex concept
Multiple indicators usually represent complex concepts more completely
Time Dimension of Research
Cross-Sectional Study
Collects data at one point in time
Provides a snapshot but cannot directly show individual change over time
Longitudinal Study
Collects or compares data across multiple times
Better for studying change, development, and sequence
Trend Study
Studies changes in a population using different samples over time
Cohort Study
Studies a category of people who share an experience or starting point
Panel Study
Studies the same individuals at multiple times
Research Proposal
Explains what will be studied, why it matters, and how the study will be conducted
Typically includes problem, literature review, purpose, questions or hypotheses, methods, ethics, and analysis plan
Allows others to evaluate feasibility and quality before data collection begins
Scientific Inquiry
Systematically investigates questions
Seeks reliable and unbiased results
Logic + Observation
Does my idea make sense?
Does the evidence support it?
Empirical Evidence
What does the evidence show?
Follow a Research Plan
Protects Against Human Error
Measurement
Careful and deliberate observation
Describes objects or events through attributes of variables
More than one measurement may represent a variable
Measurement choices should fit the research purpose
Complex concepts often benefit from multiple indicators
Variable
Characteristic that can take more than one value
Examples: age, trust level, political affiliation, or communication satisfaction
Attribute
Specific value or category of a variable
Example: married is an attribute of the variable marital status
Levels of Measurement
Nominal
Categories with no meaningful order
Examples: religion, major, or communication channel
Analyze with frequencies, mode, chi-square, or other categorical tests
Ordinal
Ordered categories with unequal or unknown distance between ranks
Examples: class rank, satisfaction level, or Likert-type response
Median and rank-based nonparametric tests are often appropriate
Interval
Equal distance between values but no true zero
Example: temperature in Fahrenheit or Celsius
Differences are meaningful but ratios are not
Ratio
Equal intervals plus a meaningful true zero
Examples: age, time, income, or number of messages
All arithmetic operations are meaningful
Memory Aid
Nominal = name
Ordinal = order
Interval = equal intervals
Ratio = real zero
Quality of Measurement
Reliability
Consistency of a measure
Test-Retest Reliability
Same measure produces similar results at different times
Alternate-Form Reliability
Different versions of the same measure produce similar results
Internal Consistency
Items intended to measure the same concept produce consistent responses
Split-Half Reliability
Compares results from two halves of the same measure
Item-Total Reliability
Checks whether each item agrees with the overall scale
Established measures may provide prior reliability evidence
Validity
Does the measure adequately represent the intended concept?
Content Validity
Measure covers the full meaning and important dimensions of the concept
Face validity
Measure appears reasonable on its surface
Expert-panel review
Experts judge whether items adequately cover the concept
Multidimensional or statistical evidence
Criterion Validity
Measure agrees with or predicts a meaningful outside criterion
Predictive validity
Measure accurately predicts a future outcome
Concurrent validity
Measure agrees with a current established measure or outcome
Construct Validity
Evidence that a measure behaves as the underlying theory predicts
Convergent validity
Related measures of the same or similar concepts should be associated
Discriminant validity
Measures of different concepts should not be too strongly associated
Representational Validity
Extent to which the operational measure faithfully represents the concept
Reliability and Validity Are Different
A consistent measure may still fail to measure the intended concept
Reliability asks whether results are consistent
Validity asks whether the correct concept is being measured
A measure generally must be reliable before it can be valid
Greater specificity can reduce some richness of a broad concept
Theory
Guides the Research Process
Helps researchers know where to look
Helps make sense of patterns
Directs future research
Helps avoid being misled by random findings
Theory + Observation
Researchers move between ideas and evidence
Deductive Reasoning
Theory -> Observation
Start with a theory
Develop a hypothesis or prediction
Gather observations/data
Determine whether evidence supports the prediction
Inductive Reasoning
Observation -> Theory
Start with observations
Identify patterns
Develop an explanation
Build toward theory
Deduction + Induction Work Together
Not completely separate approaches
Research can move back and forth between them
Evidence and Theory
Evidence can support a theory
Evidence does not absolutely prove a theory
Other theories may explain the same findings
Scientific conclusions can be probabilistic
Research Paradigms
Paradigm
Broad worldview about reality, knowledge, values, and how research should be conducted
Positivist Paradigm
Cause and Effect / Measurement
Objective reality
Variables and measurement
Hypothesis testing
Connects strongly to Deductive Reasoning
Connects strongly to Quantitative Research
Interpretive Paradigm
What does this experience mean to you?
Meaning and subjective experience
Context matters
Connects strongly to Inductive Reasoning
Connects strongly to Qualitative Research
Critical Paradigm
Who has power?
Examines power
Inequality
Challenges assumptions
Often seeks emancipation or social change, not only description
Systems Paradigm
How does everything work together?
Interconnected parts
Relationships and patterns
Changing one part of a system may affect the entire system
Ethical Research
Voluntary Participation
Participants choose whether to participate
Participation should not be forced
No Harm
Researchers must minimize potential harm to participants
Informed Consent
Participants voluntarily agree to participate
Participants understand possible risks
Participants understand what participation involves
Privacy
Protect personal information
Participant controls access to personal space, information, and circumstances
Anonymity
Researcher cannot connect responses to participant identity
Confidentiality
Researcher may know participant identity
Researcher protects participant information
Anonymity Versus Confidentiality
Anonymous = identity is not known to the researcher
Confidential = identity may be known but is not disclosed
Deception
Raises ethical concerns
Must be justified when used
Debriefing
Participants may need an explanation after participation
Institutional Review Board (IRB)
Reviews research involving human participants
Protects participants' rights and interests
Minimizes participant risk
Reviews informed consent
Reviews recruitment
Reviews confidentiality and privacy
Considers emotional, psychological, legal, social, financial, and physical risks
Approval does not remove the researcher's ongoing ethical responsibility
Plagiarism
Direct quotation without proper quotation/citation
Presenting another person's work as your own
Presenting another person's ideas as your own
Ethical Research Design
Scientific value does not automatically justify a research method
Ethics must be considered while designing the study
Good research must be scientifically sound AND ethically sound
Sampling
Why Sampling Is Necessary
Researchers cannot observe everything relevant to a phenomenon
Population
Total group of interest
Sample
Subset observed to make inferences about the population
Representativeness
How well the sample reflects the population
A large sample can still be unrepresentative
Sampling-frame quality affects representativeness
Generalizability
Extends findings from observed cases to unobserved people or events
Connected to external validity
External Validity
Extent to which findings may apply beyond the specific study sample and setting
Probability Sampling
Random selection with known, nonzero probabilities of selection
Reduces conscious and unconscious selection bias
Allows researchers to estimate sampling error
Simple Random Sampling
Every element is selected by a random process
Each element has an equal and independent chance of selection
Systematic Sampling
Select every kth element after choosing a random starting point
Risk: periodic patterns in the sampling frame can bias the sample
Stratified Sampling
Divide population into meaningful subgroups and randomly sample within each
Ensures important subgroups are represented
Multistage Cluster Sampling
Randomly select naturally occurring groups, then sample within selected groups
Useful when a complete list of individuals is unavailable or geographically scattered
EPSEM
Equal probability of selection method
Sampling Frame
List or quasi-list used to select the sample
Incomplete or outdated frames can create bias
Sampling Error
A sample never perfectly represents its population
Probability sampling makes expected error estimable
Nonprobability Sampling
Not based on random selection
Useful when probability sampling is impossible or inappropriate
Available / Accidental Participants
Select people who are easiest to reach
Fast and convenient but vulnerable to selection bias
Purposive / Judgmental Sampling
Researcher uses judgment to select participants
Chooses information-rich cases that fit the research purpose
Snowball Sampling
Participants identify other possible participants
Useful for hidden or difficult-to-reach populations
May overrepresent connected social networks
Quota Sampling
Sample reflects selected population characteristics
Sets target numbers for categories but does not randomly select within them
Probability Versus Nonprobability
Probability sampling supports statistical generalization and estimates sampling error
Nonprobability sampling supports access, depth, or information richness
Random selection is not the same as random assignment
Statistical Analysis
Choosing a Statistical Test
Depends on the research question
Depends on the level and type of data
Depends on whether groups are independent or related
Depends on the number of groups or conditions
Depends on whether statistical assumptions are satisfied
Ask First
Am I describing data, testing differences, or testing relationships?
Is the outcome categorical, ranked, or quantitative?
Are the observations independent or paired/repeated?
Parametric Statistics
Used primarily with quantitative interval- or ratio-level data
Make assumptions about population score distributions
Often compare means or examine relationships
Usually provide greater statistical power when assumptions are satisfied
Descriptive Information
Central Tendency
Mean = arithmetic average; sensitive to extreme scores
Median = middle score; useful for skewed or ordinal data
Mode = most frequent score or category
Distribution / Variability
Range = highest value minus lowest value
Variance = average squared distance from the mean
Standard deviation = typical distance of scores from the mean
Small standard deviation = scores cluster near the mean
Large standard deviation = scores are more spread out
Normal Distribution
Symmetrical bell-shaped distribution
Mean, median, and mode are approximately equal
Many parametric tests assume normality or approximate normality
Tests of Differences
z Test
Tests differences involving means or proportions
Often used when population variability is known or samples are large
t Test
Compares means when population variability is estimated from the sample
One-sample t Test
Compares one sample mean with a known or hypothesized value
Independent-groups t Test
Compares means from two separate, unrelated groups
Example: treatment group versus control group
Paired-samples t Test
Compares two related measurements
Example: the same participants before and after treatment
Analysis of Variance (ANOVA)
Compares three or more means
Tests whether at least one mean differs from the others
May use planned comparisons or post hoc tests
Post hoc tests identify which groups differ after a significant overall result
ANCOVA
Compares group means while statistically controlling for another quantitative variable
MANOVA
Compares groups on two or more related outcome variables simultaneously
MANCOVA
MANOVA while statistically controlling for one or more covariates
Tests of Relationships
Correlation
Measures the direction and strength of association between two quantitative variables
r ranges from -1 to +1
Positive = variables move in the same direction
Negative = variables move in opposite directions
Near zero = little or no linear relationship
Strength depends on distance from zero, not the positive or negative sign
Closer to -1 or +1 = stronger relationship
Correlation does not prove causation
Pearson Product-Moment Correlation
Symbol: r
Best for two continuous numerical variables with an approximately linear relationship
Parametric test; assumptions should be checked
Report r, degrees of freedom, p value, and supporting means and standard deviations
Degrees of freedom for a basic Pearson correlation: df = n - 2
Example: 26 participants gives df = 24
Spearman Rank-Order Correlation
Also called Spearman's rho or rank-order correlation
Symbol: rs
Best for ordinal or ranked data or when Pearson assumptions are not met
Nonparametric alternative to Pearson correlation
Report rs, p value, and degrees of freedom or sample size
Pearson Versus Spearman
Pearson = continuous numerical data and linear relationship
Spearman = ranked or ordinal data or violated Pearson assumptions
Both describe direction and strength and require the test to be named clearly
Partial Correlation
Examines the relationship between two variables while statistically controlling a third variable
Asks whether the relationship remains after the third variable's influence is held constant
Report partial correlation value, p value, and degrees of freedom
Shared Variance
r-squared estimates the proportion of variance shared by two variables
Calculate by squaring the correlation coefficient
Example: .86 x .86 = .74, or about 74 percent shared variance
Squaring removes the negative sign; correlations of +.50 and -.50 share the same amount of variance
How to Read a Correlation Result
Identify Pearson r, Spearman rs, or partial correlation
Check the sign for direction
Check the absolute size for strength
Check the p value for statistical significance
Interpret the result using the actual variables and research context
Do not make a causal claim from correlation alone
Regression
Uses one predictor to estimate or explain a quantitative outcome
Multiple Regression
Uses two or more predictors to estimate a quantitative outcome
Shows each predictor's unique contribution while controlling for the others
Logistic Regression
Predicts a categorical outcome, often one with two categories
Measures of Association
r = strength and direction of a linear relationship
R-squared = proportion of outcome variation explained by the model
Beta weight = direction and relative contribution of a predictor in regression
Nonparametric Statistics
Used with nominal, ordinal, ranked, frequency, or categorical data
Make fewer population-distribution assumptions
Often compare frequencies, categories, ranks, or medians
Useful when parametric assumptions are violated or samples are small
Results are often clarified with tables
Chi-Square Tests
Used with categorical frequencies
Goodness-of-Fit Chi-Square
Compares observed frequencies with expected frequencies
Uses one categorical variable
Chi-Square Test of Independence
Tests whether two categorical variables are associated
Uses a contingency table
Yates's correction may be used for continuity
Fisher's exact test may be used with small expected cell frequencies
Wilcoxon Rank Sum / Mann-Whitney U
Compares two independent groups using ranks
Alternative to independent-groups t test
Kruskal-Wallis
Compares three or more independent groups using ranks
Alternative to one-way independent-groups ANOVA
McNemar Test
Compares two related categorical measurements
Example: paired yes/no responses before and after an intervention
Cochran's Q
Compares three or more related categorical measurements
Wilcoxon Signed-Rank
Compares two related or matched measurements using ranks
Alternative to paired-samples t test
Friedman ANOVA by Ranks
Compares three or more related conditions using ranks
Alternative to repeated-measures ANOVA
Parametric and Nonparametric Connections
Independent-groups t test <-> Mann-Whitney U / Wilcoxon rank sum
Paired-samples t test <-> Wilcoxon signed-rank
One-way independent-groups ANOVA <-> Kruskal-Wallis
Repeated-measures ANOVA <-> Friedman test
Statistical Significance and Practical Importance
p Value
Probability of obtaining results this extreme if the null hypothesis were true
A small p value provides evidence against the null hypothesis
Does not show the size or importance of an effect
Effect Size
Shows the magnitude or practical strength of a difference or relationship
Should be reported even when a result is not statistically significant
Sample Size Matters
Large samples can make small effects statistically significant
Small samples may miss meaningful effects because of low power
Reporting Statistical Results
Clearly identify the test used
Report the test statistic
Report degrees of freedom when applicable
Degrees of freedom (df) reflect how much independent information is available to estimate a statistic
Report number of observations (N)
Report observations per group or cell when relevant
Report significance level (p)
Report effect size
Include supporting descriptive statistics or mean ranks
Use tables when they improve comprehension
Correlation Reporting
Clearly identify Pearson, Spearman, or partial correlation
Include coefficient, direction, p value, and degrees of freedom or sample size
Include means and standard deviations for Pearson when appropriate
Report nonsignificant findings honestly, not only significant results
Statistical significance does not automatically mean a relationship is strong or practically important
Qualitative Research Methods
Chapter 13: Participant Observation
Definition and Purpose
Qualitative method conducted in a natural setting
Researcher observes and may participate in events being studied
Provides a comprehensive and detailed understanding
Reveals subtle attitudes, behaviors, and communication processes over time
Appropriate Research Topics
Practices
Routine, socially recognized forms of communication or action
Episodes
Important or dramatic events that unfold over time
Encounters
Immediate interaction between two or more people
Roles and relationships
Communication tied to social positions and connections between people
Groups and organizations
Communication patterns within small groups or formal institutions
Settlements, social worlds, lifestyles, and subcultures
Communication codes shared within communities or loosely bounded social groups
Ethnography of Communication
Studies communication codes within a speech community
Speaking is structured, distinctive, and social
SPEAKING Framework
Situation
Physical and social setting of the communication
Participants
People involved and their relationships to one another
Ends
Goals, purposes, and outcomes
Act characteristics
Form and content of what is said or done
Key
Tone or manner, such as serious, playful, or sarcastic
Instrumentalities
Channel, language, dialect, or communication medium
Norms of interaction
Rules for producing and interpreting communication
Genres
Recognized type of communication, such as joke, prayer, lecture, or story
Critical Ethnography
Examines power relationships
Gives voice to groups with less power
Uses research to support social change
Case-Study Research
Studies a specific, unique, bounded system
Single-case study examines one case deeply
Collective-case study compares several cases
Within-case analysis examines each case separately
Cross-case analysis identifies themes across cases
Choreography of Participant Observation
Warm-Up Period
Develop a broad guiding question
Review literature to build theoretical sensitivity
Identify researcher biases through self-reflection
Choose an information-rich site through purposive sampling
Select the researcher's role
Gain access through gatekeepers and sponsors
Plan for ethics and sufficient time in the field
Floor Exercise Period
Sample information-rich participants, activities, and scenes
Use maximum-variation, typical-case, snowball, theoretical, or critical-case sampling
Maximum variation = deliberately include diverse cases
Typical case = select a case that represents what is ordinary
Theoretical sampling = choose later cases to develop emerging ideas
Critical case = select a case that dramatically illustrates the phenomenon
Create field notes from jottings and expanded notes
Record description, preliminary analysis, and reflexive observations
Use photographs, film, or video when appropriate
Make ethical decisions as unexpected situations arise
Triangulate data, methods, researchers, and theories
Data triangulation = compare information from different people, settings, or times
Method triangulation = examine the issue with more than one method
Researcher triangulation = compare interpretations from multiple researchers
Theory triangulation = interpret findings through more than one theoretical lens
Member checking = ask participants whether interpretations ring true
Remain flexible as the study develops
Cool-Down Period
Reach saturation when additional data provide no new insights
Withdraw gradually from the research site
Complete a more comprehensive analysis
Return for additional data or member checking when necessary
Researcher Roles
Complete participant
Participates fully while people do not know research is occurring
High access but serious deception and consent concerns
Participant as observer
Participates actively and participants know the research role
Observer as participant
Primarily observes with limited interaction; participants know the research role
Complete observer
Observes without participating in the group's activities
Roles vary by participant awareness and researcher involvement
Field Notes and Reflexivity
Field journal is the backbone of participant observation
Jottings are expanded into detailed chronological notes
Attend to who, what, when, where, and why
Description records what happened
Preliminary analysis records possible patterns and meanings
Reflexive notes record the researcher's choices, feelings, and influence
Reflexive journal documents decisions, feelings, and possible biases
Creates an audit trail that supports dependability
Strengths
Depth and rich detail
Natural context
Flexibility and adaptability
Useful for processes that unfold over time
Limitations
Cannot provide statistical descriptions of large populations
Findings depend heavily on the researcher's integrity and interpretation
Can require substantial time in the field
Chapter 14: Qualitative Interviewing
Definition and Purpose
Conversation with a purpose
Uses open-ended talk to understand an informant's meanings and experiences
Flexible, iterative, and continuously revised
Usually semistructured or unstructured
Best for questions about how people experience, interpret, or give meaning to something
Participant accounts are interpretations, not neutral recordings of facts
Appropriate Research Purposes
Learn about experiences that cannot be directly observed
Understand thoughts and feelings in rich detail
Study participants' language, vocabulary, and idioms
Triangulate findings and conduct member checking
Study the interview itself as a communicative performance
Qualitative Interview as a Speech Event
Has a beginning, middle, and end
Is jointly created by interviewer and informant
Resembles conversation but has a research purpose
Interviewer encourages the informant to do most of the talking
Interviewer's wording, identity, reactions, and rapport can shape the answers
Planning the Interview
Identify the research purpose
Select appropriate informants
Develop an interview protocol as a guide rather than a rigid script
Include an introduction, consent information, questions, probes, and closing
Interview Protocol Types
Structured
Same wording and order for every participant
Improves consistency but limits depth and flexibility
Semistructured
Prepared questions with flexible wording, order, and probes
Balances consistency across participants with depth
Unstructured
Broad topics guide an emergent conversation
Maximizes participant direction but makes interviews harder to compare
Open-Ended Question Types
Descriptive questions explore experiences and examples
Example: Can you walk me through what happened?
Structural questions reveal how participants organize meanings or categories
Example: What different kinds of support did you receive?
Contrast questions uncover similarities and differences
Example: How was that experience different from the earlier one?
Probes
Follow-up questions that request detail, clarification, meaning, or examples
Should emerge from the participant's answer rather than sound mechanical
Executing the Interview
Build rapport through trust and respect
Listen carefully and avoid dominating the conversation
Avoid leading questions and putting words in the participant's mouth
Probe for clarification, examples, and greater detail
Adapt questions to what the informant says
Record interviews with permission and create field notes
Genres of Qualitative Interviewing
Ethnographic conversation
Informal interview that occurs naturally during fieldwork
Depth interview
Extended one-on-one interview exploring a participant's viewpoint in detail
Group or focus interview
Usually includes approximately 7-12 relevant participants
Unit of analysis is the group discussion
Group interaction can reveal agreement, disagreement, norms, and shared language
Not ideal when privacy, hierarchy, or peer pressure may silence participants
Multiple groups are used until saturation
Narrative interview
Elicits stories participants use to make sense of their lives
May focus on life history, oral history, turning points, or a specific event
Postmodern interview
Challenges the idea that the researcher can present one complete objective account
Preserves multiple voices and recognizes that interviewer and participant create the account together
Trustworthiness
Dependability: maintain field notes and a reflexive audit trail
Can another person track how the study and analysis developed?
Confirmability: support conclusions with detailed evidence and triangulation
Are interpretations grounded in the data rather than unsupported researcher preference?
Credibility: check whether findings ring true to participants
Qualitative counterpart to asking whether findings are believable and accurate
Transferability: provide enough detail for readers to judge relevance elsewhere
Researcher supplies thick description; reader judges whether findings fit another context
Saturation determines when enough interviews have been completed
Reached when additional interviews repeat known patterns and add no meaningful insight
Chapter 15: Social Text Analysis
Definition of a Social Text
Naturally occurring symbols in use
Different from researcher-generated field notes or interview transcripts
May include conversations, letters, diaries, photographs, films, speeches, memorials, advertisements, or websites
Uses of Social Texts
May serve as the primary data
May triangulate participant observation and interviews
Multiple texts can corroborate and contextualize one another
Work Versus Text
Work is the original communication message or event
Text is the researcher's interpretation of that work
Textualization constructs a reasonable interpretation grounded in the work
One work can support multiple credible texts or interpretations
Interpretation must be insightful and remain grounded in evidence from the work
Documents as Social Texts
Create a paper trail of events, policies, values, and processes
Can reveal information unavailable through direct observation or interviews
Must be checked for accuracy, bias, representativeness, and missing voices
Meaning depends on authorship, purpose, history, use, and context
Enacted Talk and Transcription
Transcription transforms recorded talk into analyzable text
No transcript is completely neutral or free of theory
What is transcribed depends on the research purpose
Category design should be discriminable, exhaustive, and contrastive
Discriminable = categories can be clearly distinguished
Exhaustive = categories cover all relevant possibilities
Contrastive = differences between categories serve the research purpose
Readability makes the transcript understandable
Tractability makes the system practical for the researcher to use
Approaches to Social Text Analysis
Communication Criticism
Analyzes public texts such as speeches, television programs, and films
Moves from textualization to analysis to interpretation
Discourse Analysis
Studies the structure and functions of language in use
Structural approach asks how discourse is organized
Functional approach asks what discourse accomplishes
Conversation Analysis
Examines interaction at a microscopic level
May study turn taking, pauses, overlap, repair, and sequence
Narrative Approach
Studies stories for their form, content, and function
Form = how the story is organized or told
Content = themes, events, characters, and meanings
Function = what the story accomplishes for the teller or audience
Performative or Dramatistic Approach
Uses performance or studies communicative performances to understand meaning
Treats the voice and body as tools for understanding enacted meaning
Semiotic Approach
Studies how signs communicate socially constructed meanings in context
Meaning is not naturally contained in a sign; communities learn and share the association
Analytic Frameworks
Provide a lens for interpreting a text
Should guide the researcher without becoming blinders
Can be combined when appropriate for the research question
Strengths
Provides close access to naturally occurring communication
Captures meanings that numerical frequency alone cannot explain
Supports triangulation with other qualitative methods
Limitations
Meaning can be misunderstood when a text is separated from its context
Analysis and transcription are labor intensive
Interpretation lacks the numerical precision of quantitative text analysis
Qualitative Versus Quantitative Text Analysis
Qualitative analysis interprets meanings, uses, patterns, and context
Quantitative content analysis counts coded categories and tests numerical relationships
The approaches can complement one another in a mixed-method study
Connections Across Chapters 13-15
Participant observation studies communication where it naturally occurs
Qualitative interviewing asks participants to explain experiences in their own words
Social text analysis interprets the messages and symbols people naturally produce
All three methods are qualitative, flexible, interpretive, and context dependent
All three rely on the researcher as an instrument of interpretation
All three use purposive sampling and seek saturation rather than statistical generalization
All three are evaluated through dependability, confirmability, credibility, and transferability
Using the methods together strengthens findings through triangulation
Quick Method Selection Guide
Use a Survey When
You need standardized answers from many people
You want estimates, comparisons, or relationships that may generalize to a population
Use an Experiment When
You need to test whether changing one variable causes a change in another
Manipulation, control, and random assignment are possible and ethical
Use Participant Observation When
You need to see communication in its natural setting
Context, behavior, group norms, and processes over time matter
Use Qualitative Interviews When
You need detailed personal meanings, experiences, or explanations
The phenomenon is private, sensitive, or cannot be directly observed
Use Focus Groups When
Group interaction, shared language, agreement, or disagreement is part of the data
The topic is safe enough for participants to discuss together
Use Social Text Analysis When
The research question concerns the meaning or use of existing messages and symbols
Documents, conversations, images, speeches, media, or online content are the evidence
Choose Quantitative Methods When
The goal is measurement, comparison, prediction, or testing relationships
Choose Qualitative Methods When
The goal is depth, context, meaning, process, or lived experience
Choose Mixed Methods When
One method alone cannot answer every part of the research problem
Quantitative patterns and qualitative explanations are both needed
]