Plato's Allegory of the Cave

Prisoners chained in a cave, seeing only shadows on a wall and believing them to be reality. The physical world we perceive is like shadows on a cave wall, an imperfect reflection of true reality.

a way of learning and knowing about the world

experiential reality and agreement reality

the things you know as a function of your direct experience

things you consider real because you’ve been told they’re real, and everyone else seems to agree they are real

the process of thinking, forming judgments, and drawing conclusions, often using both logical and probabilistic approaches.

Prone to publication bias, uneven study quality, and inconsistent measures. Example: A meta-analysis on social media use is weakened when studies define the construct differently.

the process of determining a relationship between a cause and its effect, essentially asking "how" or "why" something happened

probabilistic reasoning

a framework for making decisions and drawing conclusions under conditions of uncertainty, relying on probability theory rather than absolute certainty

can both inform and constrain human reasoning by providing a shared conceptual framework and a basis for establishing standards of justification

reasoning from authority, a logical fallacy and a practical method of acquiring knowledge where an argument is supported by the opinion of a credible source

errors in personal human inquiry

inaccurate observation overgeneralization selective observation made-up information ex post facto hypothesizing illogical reasoning ego involvement in understanding premature closure of inquiry mystification

drawing general, negative conclusions from a single experience or limited evidence, often using words like "always" or "never"

selective observation

when we see only those patterns that we want to see or when we assume that only the patterns we have experienced directly exist

broadly defined as false or inaccurate information

ex post facto hypothesizing

creating a hypothesis that explains results after they have already been observed

thinking or arguing with flawed logic, invalid premises, or insufficient evidence, which leads to conclusions that are unsound, false, or incoherent

ego involvement in understanding

state where understanding and learning are influenced by personal stakes, often involving a concern for how one's self-image, status, or competence is perceived by others

premature closure of inquiry

a bias or error where an investigation or assessment is halted prematurely because a plausible explanation or solution is found, leading to incomplete data analysis or a misdiagnosis

how reality is distorted, masked, or obscured through symbolic representations or cognitive frameworks

the systematic effort to secure answers to questions

research completed to learn about relationships among variables, regardless of any immediate commercial product or service

research completed to develop a product or solve an immediate practical problem

types of social research

descriptive exploratory explanatory

observe and describe the characteristics of a population, phenomenon, or situation without manipulating variables or establishing cause-and-effect relationships

a flexible research approach for thoroughly investigating poorly defined problems, understanding scope, and generating initial hypotheses to guide future, more structured studies

a research method that seeks to identify and understand the causes and effects between variables, answering "why" and "how" questions about a phenomenon

overall scientific enterprise

the entire system of human activities aimed at generating and applying scientific knowledge

an inquiry is a systematic process of investigating questions through evidence collection to reach informed conclusions, while a problem is a situation or obstacle that needs to be overcome or a goal to be achieved

a formal explanation for observed phenomena, based on extensive evidence, facts, and tested hypotheses

research plan, sampling, observation, and data processing

findings reported in a study, while "statistics" is the field and methodology for collecting, analyzing, and interpreting data to understand patterns and make decisions

author interprets, analyzes, and explains the significance of the findings in relation to the research questions and existing literature

the imparting or exchanging of information or news

a set of verbal and nonverbal cues that communicators exchange

the words people use in communication

communication elements beyond the words themselves

methods that described observations in predominantly non-numerical terms

inquiries in which observations are expressed predominantly in numerical terms

a characteristic or property that can vary (assigned either values or attributes)

a characteristic or quality that describes an object

a characteristic to which only one number may be assigned, does not change

An expectation about events based on generalizations of the assumed relationship between variables

the variable that is hypothesized to cause, or lead to, variation in another variable

the variable whose variation is hypothesized to depend on or be influenced by the other variable

combines quantitative results from many studies to reveal overall sizes of effect that exist among variables

the logical structure of relationships, calculus refers to any deductive system

Help

Options

focusNode

Didn't know it?
click below

Knew it?
click below

Don't Know

Remaining cards (0)

Know

retry

shuffle

restart

0:00

Embed Code - If you would like this activity on your web page, copy the script below and paste it into your web page.

Normal Size Small Size show me how

COMM 561 Final

Term	Definition
Plato's Allegory of the Cave	Prisoners chained in a cave, seeing only shadows on a wall and believing them to be reality. The physical world we perceive is like shadows on a cave wall, an imperfect reflection of true reality.
science	a way of learning and knowing about the world
two realities	experiential reality and agreement reality
experiential reality	the things you know as a function of your direct experience
agreement reality	things you consider real because you’ve been told they’re real, and everyone else seems to agree they are real
human reasoning	the process of thinking, forming judgments, and drawing conclusions, often using both logical and probabilistic approaches.
native human inquiry	Prone to publication bias, uneven study quality, and inconsistent measures. Example: A meta-analysis on social media use is weakened when studies define the construct differently.
causal reasoning	the process of determining a relationship between a cause and its effect, essentially asking "how" or "why" something happened
probabilistic reasoning	a framework for making decisions and drawing conclusions under conditions of uncertainty, relying on probability theory rather than absolute certainty
tradition	can both inform and constrain human reasoning by providing a shared conceptual framework and a basis for establishing standards of justification
authority	reasoning from authority, a logical fallacy and a practical method of acquiring knowledge where an argument is supported by the opinion of a credible source
errors in personal human inquiry	inaccurate observation overgeneralization selective observation made-up information ex post facto hypothesizing illogical reasoning ego involvement in understanding premature closure of inquiry mystification
inaccurate observation	an inaccurate observation
overgeneralization	drawing general, negative conclusions from a single experience or limited evidence, often using words like "always" or "never"
selective observation	when we see only those patterns that we want to see or when we assume that only the patterns we have experienced directly exist
made-up information	broadly defined as false or inaccurate information
ex post facto hypothesizing	creating a hypothesis that explains results after they have already been observed
illogical reasoning	thinking or arguing with flawed logic, invalid premises, or insufficient evidence, which leads to conclusions that are unsound, false, or incoherent
ego involvement in understanding	state where understanding and learning are influenced by personal stakes, often involving a concern for how one's self-image, status, or competence is perceived by others
premature closure of inquiry	a bias or error where an investigation or assessment is halted prematurely because a plausible explanation or solution is found, leading to incomplete data analysis or a misdiagnosis
mystification	how reality is distorted, masked, or obscured through symbolic representations or cognitive frameworks
research	the systematic effort to secure answers to questions
basic research	research completed to learn about relationships among variables, regardless of any immediate commercial product or service
applied research	research completed to develop a product or solve an immediate practical problem
types of social research	descriptive exploratory explanatory
descriptive	observe and describe the characteristics of a population, phenomenon, or situation without manipulating variables or establishing cause-and-effect relationships
exploratory	a flexible research approach for thoroughly investigating poorly defined problems, understanding scope, and generating initial hypotheses to guide future, more structured studies
explanatory	a research method that seeks to identify and understand the causes and effects between variables, answering "why" and "how" questions about a phenomenon
overall scientific enterprise	the entire system of human activities aimed at generating and applying scientific knowledge
problem or inquiry	an inquiry is a systematic process of investigating questions through evidence collection to reach informed conclusions, while a problem is a situation or obstacle that needs to be overcome or a goal to be achieved
theory	a formal explanation for observed phenomena, based on extensive evidence, facts, and tested hypotheses
research methods	research plan, sampling, observation, and data processing
results/statistics	findings reported in a study, while "statistics" is the field and methodology for collecting, analyzing, and interpreting data to understand patterns and make decisions
discussion	author interprets, analyzes, and explains the significance of the findings in relation to the research questions and existing literature
communication	the imparting or exchanging of information or news
message	a set of verbal and nonverbal cues that communicators exchange
verbal cues	the words people use in communication
nonverbal cues	communication elements beyond the words themselves
qualitative methods	methods that described observations in predominantly non-numerical terms
quantitative methods	inquiries in which observations are expressed predominantly in numerical terms
variable	a characteristic or property that can vary (assigned either values or attributes)
attribute	a characteristic or quality that describes an object
constant	a characteristic to which only one number may be assigned, does not change
hypothesis	An expectation about events based on generalizations of the assumed relationship between variables
nondirectional hypothesis	H₁: µ₁ ≠ µ₂
directional hypothesis	H₁: μ₁ > μ₂
null hypothesis	H0: μ1 = μ2
independent variable	the variable that is hypothesized to cause, or lead to, variation in another variable
dependent variable	the variable whose variation is hypothesized to depend on or be influenced by the other variable
meta-analysis	combines quantitative results from many studies to reveal overall sizes of effect that exist among variables
abstract calculus	the logical structure of relationships, calculus refers to any deductive system
conditional syllogism	a logical argument with two premises and a conclusion, where the major premise is an "if-then" statement
constructs	a complex, synthesized idea, theory, or concept formed from simpler elements
rules of correspondence	define the connection between a theoretical model and the observable world
description	detailed account, either in words or pictures, that presents the characteristics and features of a person, place, thing, or event
explanation	a statement of the facts or reasons that make something clear and understandable
prediction	a statement about what is expected to happen in the future, often based on existing knowledge, data, or experience
control	an element that remains unchanged or unaffected by other variables
conceptual definitions	focus on the broader understanding of a concept
operational definitions	specifies how a concept, variable, or construct will be measured in a scientific study
manipulated independent variables	"Stimulus Variables" - Introduced	Controlled
measured/assigned variables	Not introduced/controlled, but are carefully observed and measured
direct classification	Variables operationally defined concepts by simple identification or classification of observable characteristics of information
intersubjectivity	The degree to which different researches with essentially different beliefs draw essentially the same interpretations of the meaning of observations
attenuation	wh en relationships in research get lost in the background noise
levels of measurement	See below
nominal	categories with no inherent order and no inherent quantitative properties (e.g., eye color)
ordinal	categories with a meaningful order but inconsistent spacing (e.g., satisfaction surveys)
interval	have order and equal spacing but no true zero point (e.g., IQ scores)
ratio	possess all these qualities (order, equal intervals) plus a true zero (e.g., height or weight)
categorical data	represents characteristics or qualities of a subject that can be grouped into distinct categories, like "male" or "female," rather than quantities
continuous variables	variable that can take on any value within a given range, rather than being restricted to a fixed number of distinct values (e.g. height of a plant)
"mediator	it explains the relationship between the independent variable (IV) and the the mechanism for why the IV brings about the DV
moderator	a qualitative (e.g. sex, race, class) or quantitative (e.g., level of reward) variable that affects the direction and/or strength of the relation between an independent or predictor variable and a dependent or criterion variable
crucial experiment	test designed to decide between two or more competing theories or hypotheses by producing a result that strongly supports one theory and makes the others improbable
exogenous variables	a variable in a causal model or causal system whose value is independent from the states of other variables in the system
endogenous variables	a variable in a causal model or causal system whose value is determined by the states of other variables in the system.
deductive	a form of argument in which a valid conclusion necessarily follows from premises, uses syllogisms
inductive	the process by which we conclude what is true of certain individuals is true of a class, what is true of part is true of the whole class, or what is true at certain times will be true in similar circumstances at all times
advantages/ disadvantages of the approaches	Deductive reasoning is efficient but passive and flawed if premises are wrong. Inductive reasoning builds deeper, active learning but is slower and risks incorrect generalizations.
falsification of a theory	theories must be capable of being falsified if they are incorrect
tentativeness of a theory	theories must be tentative - demands scholars recognize a theory's answers are provisional
reliability	the consistency and dependability of a measurement or instrument
false precision	when a researcher claims more precision in analyzing data then was present in the original data
fallacy of misplaced precision	when numerical data is presented with a level of exactness that is not justified by the actual data or situation, giving the false impression of accuracy and certainty
reliability coefficient	a statistical index, usually ranging from 0 to 1, that indicates the consistency of a measurement or test by estimating the proportion of true score variance to total variance
test-retest reliability	giving the measurement twice and reporting consistency between scores
alternate forms (parallel forms) reliability	constructing different forms of the same test from a common pool of measurement items, giving each of the different forms to the SAME group of people and determining the degree of consistency between them
split-half reliability	divides a test in 2 parts, scores them separately, and checks consistency between the 2 scores (sometimes called the odd-even reliability)
item to total reliability	computes the correlation of items with the total test (e.g. answering first question right on the exam might mean you do better than someone who answers first question wrong)
intercoder reliability	measures the extent to which two or more independent coders agree on the codes applied to the same data, ensuring the consistency and trustworthiness of qualitative analysis
validity	the degree to which a test, measurement, or study accurately measures what it is intended to measure
face validity	involves researchers looking at the content of the measurement items and advancing an argument that on its face the measure seems to identify what it claims
expert jury validity	involves having a group of experts in the subject matter examine the measurement and judge its merits
criterion validity	consists of methods that assess a measures worth by examining its relation to some outside criterion.
concurrent validity	correlates a new measure with a previously validated measure of the same thing.
predictive validity	the degree to which a measure predicts known groups in which the constructs must exist
construct validity	requires that a new measure be administered to subject along with at least 2 other measures (one should be a valid measure of a construct that is known conceptually to be related to the new measure)"
Cronbach's coefficient alpha	when you want to see consistency with which people react to items on a scale... AND when there is no "correct" or "incorrect" answers.
Thurstone equal appearing interval scales	composed of statements related to some topic issues: time consuming, tedious, and expensive
Likert scales	presents statements that reflect clear positions on an issue then asks subjects to indicate their response on a 5-point scale
Guttman scalogram	involves a series of statements dealing with one topic and arranged according to level of intensity
semantic differential-type scales	involves using parts of a "bipolar" adjectives that could be used to indicate their relationships
premise	a statement in a logical argument that is the foundation for others drawn from it
subject tracings	keywords that list other subject headings under which the book is also listed.
keywords	terms under which information about the topic may be found
argument	a process of advancing conclusions based on reasons	evidence
Einstein syndrome	when inexperienced researchers believe they can ignore past work and use entirely new ideas and methods
annotations	brief comments of your own about the work... additional explanations, comments, evaluations, or criticisms to help understand material"
plagiarism	the false assumption of authorship toward the wrongful act of taking the product of another person's mind, and presenting it as one's own
search engines	finding sources on the internet, it can rummage through internet resources for key words and content areas
exhaustive literature review	research surveys that include all material related to the subject
exemplary literature review	surveys the most important contributions
explication	requires you to complete a lit review that makes an issue clear and comprehensive
internal organizers	making active use of phrases that preview, summarize, and transitions between main points... can underscore your content
heuristic	a shortcut or rule of thumb for problem-solving
research prospectus	a complete proposal for a research activity to be completed at future date
ecological fallacy	when a researcher uses data from groups of people to draw conclusion about individuals
the Delphi fallacy	use of vague predictions as research claims
the Jeanne Dixon fallacy	making multiple predictions and claiming partial support
the Patchwork quilt fallacy	making no predictions but offering explanations after the fact
evidence	the information that scholars use to support claims
factual evidence	consists of descriptions and characterizations of things (events of past, new samples, etc.)
opinions	interpretations of the meaning of collections of facts
reports	accounts of what took place whether by participants or by outside observers
primary sources	provide information from individuals who have first hand experience with the events reported
secondary sources	provide information obtained from individuals who do not have firsthand experience with the events
statistics	quantitative reports based on observation in a sample
parameters	numbers that describe the population
expert opinions	opinions from people who are experts in their field
argument from definition	reasoning that submits that things do or do not belong in a certain class of things
argument from example and generalization	taking some particular cases and arguing that what is true of the instances is generally true in the population of events
argument from analogy	compares two things known to be alike in one or more features and suggests that they will be alike in other features as well
literal analogy	compares something to an event or object that really exists
figurative analogy	compares something to a hypothetical situation
causal argument	reasoning that a given factor is responsible for producing certain other results
syllogism	set of 2 premises that result in a conclusion
categorical syllogism	syllogisms that starts with a categorical statement	(ex- Major premise: all men are mortal.... Minor Premise: Socrates is a man.... Conclusion: Socrates is mortal...)
disjunctive syllogism	makes an "either-or" statement in the major premise
conditional syllogism	major premise makes an "if-then" statement
nomothetic research	designed to find general law-like patterns that apply to many instances
idiographic research	designed to develop a full understanding of a particular event or individual
text	what we believe are the actual messages or words of a communicator (ex- TV shows, political debates, and influential movies)
rhetorical criticism	the use of standards of excellence to interpret and evaluate communication
rhetoric	Study of available means of persuasion
criticism	Standards of excellence are announced for application
impressionistic criticism	the statement of opinion (or personal impression) made by reviewers, such as those who "review" movies for newspapers and magazines
invention	types and sources of ideas
ethos	credibility of a speaker, "ethical appeal”
pathos	the use of emotional or motivational appeals by a speaker, “pathetic appeal”
logos	logical appeals, including evidence and reasoning used by speaker (Greek word with no translation)
narrative fidelity	the consistency of new accounts with others stories people have heard
content analysis	a systematic method to analyze the content and treatment of communication, which usually results in the development of objectives and quantitative information
coding units	categories used to count the communication forms in the examples chosen
utterance	what a person actually says in conversation
interactional analysis	interactional analysis is a research method for systematically coding and interpreting communication behaviors in a specific interaction
relational analysis	relational analysis is a broader term that examines how people's feelings and perceptions influence and are influenced by their relationships
interaction analysis	focuses on ways of tracking individuals acts of communications
relational control analysis	tracks message sequences to determine the relative patters of position and control in the relationship (statements categorized as "one-up", "one-down", and "one-across"
fieldwork	Collecting data in natural settings.	Collecting data in natural settings.
Purposes:	Study of people in their natural daily lives; used when self-reports aren’t feasible, settings are unexplored, or researchers aim to develop grounded theory.
participant observation	inquires in which the researcher takes the role of an active agent in the situation under study
naturalistic studies	nonexperimental inquires completed as subjects are involved in the natural course of their lives
full participant observation	research is characterized by the investigator's gathering data while taking part in the activities of a group - while concealing his or her research identity.
participant as observer research	the group to be studied is made aware of the researcher role
accretions	deposits of material left by some action
urban archaeology	area of study that looks at deposits
erosions	the wear or use of objects
life-history studies	biography of a person which has been obtained through interview and guided conversations
grounded theory	a set of explanations that has immediate relevance to a specific field of setting under investigation
analytic induction	when a researcher begins with some very tentative hypotheses that they may apply in fieldwork
ethnography	research in which the investigator participates, overtly or covertly, in people's lives for an extended period of time, collecting whatever data are available to describe behavior.
ethnomethodology	study of the mundane and ordinary activities of everyday life, concentrating on the methods used by people to report their commonsense practical action to others in acceptable, rational terms “the new ethnography
time-budgeting studies	inquires in which the researcher asks the subjects to keep a detailed diary over a given period of time
community studies	involves a researcher or team in studying a whole community of people, usually a small town or village, or possible a part of a larger town
case study	intense inquires about single events, people, or social unites
negative case study	researcher attempts to obtain a case that has a potential to negate a generally accepted view
content analysis	systematic method to analyze the content and treatment of communication, which usually results in the development of objectives and quantitative information.
myths of content analysis	1. Content analysis is limited to simple analyses 2. Anyone can do content analyses, it does not take any special preparation 3. The term content analysis applies to ALL examination of messages 4. Content analysis is for academic use only"
rhetorical analysis	examines how a text uses communication strategies to persuade an audience to achieve a specific purpose
narrative analysis	goal is to understand relationships between a text and social reality
discourse analysis	engages in characteristics of manifest language and word use - description of topics in media texts - through consistency and connection of words to theme analysis of content and the establishment of central terms.
structuralist or semiotic analysis	focuses on deep meanings of messages
interpretive analysis	focuses on formation of theory from the observation of messages and the coding of those messages
conversation analysis	technique for analyzing naturally occurring conversations, used by social scientists in psych, comm, and sociology
critical analysis	conducted in tradition with cultural studies, has been widely used method for analysis of media messaged
normative analysis	a prescriptive approach to issues dealing with what "should be" rather than "what is"
computers and qualitative message analysis	computer adjuncts have been developed to support the tasks of these various qualitative methods
content analysis features	objectivity-intersubjectivity a priori design reliability validity generalizability replicability hypothesis testing based on theory
objectivity-intersubjectivity	provide a description or explanation of a phenomenon in a way that avoids biases of the investigator(s)
a priori design	a plan that is conceptualized and defined using reason and theory before any data has been collected or empirical testing has occurred
reliability	the internal consistency of the methods of a content analysis
validity	the content analysis actually measures what is is claimed
generalizability	extent to which findings may be applied to other cases, usually to a larger set that is the defined population from which a study's sample has been drawn
replicability	involves repeating a study with different cases or in a different context, checking to see if similar results are obtained each time.
hypothesis testing based on theory	uses sample data to assess a theory-based claim about a population by setting up a null hypothesis (the default or opposite of the theory) and an alternative hypothesis (the actual theory to be tested)
unit of analysis	the element on which data are analyzed and for which findings are reported
manifest content	the elements that are physically present and countable
latent content	unobserved concepts that can not be measured directly but can be represented or measured by one or more indicators
substance characteristics	focuses on the physical properties (observable without chemical change, like color or density) and chemical properties (observed during chemical change, like flammability) that define and identify pure substances and mixtures
formal features	the formal, systematic process of conducting research itself, which is characterized by a structured, objective, and replicable methodology, and the analysis of "formal features" within a text or artwork, focusing on intrinsic elements like diction, image
human coding vs. computer coding	Human coding refers to tasks performed by people to communicate with and control computers, while computer coding (or programming) is the technical process of writing instructions for a computer to execute, using structured languages with strict rules of
steps of the typical process of content-analytic research	developing clear research questions, selecting relevant content, defining the units and categories for analysis, developing a coding scheme, systematically coding the content according to the scheme, and finally, analyzing the results to draw conclusions
type of content analysis	descriptive inferential psychometric clinical content analysis thematic content analysis predictive
descriptive	describe a particular message pool in almost archival fashion
inferential	where researchers wish to make conclusions about sources or receivers solely from content analysis studies
psychometric	process of evaluating psychological tests and measurements by analyzing their quantitative properties, such as item difficulty and reliability, to ensure the test is fair, accurate, and valid for measuring the intended psychological attributes like knowle
clinical content analysis	seeks to provide clinical diagnosis for an individual through analysis of messages generated by that individual
thematic content analysis	to measure a psychosocial trait or state through message analysis
predictive	the prediction of some outcome or effect of the messages under examination
"unit	"1. Serves as the basis for identifying the population and drawing a sample 2. The component on which variables are measured 3. Serves as the basis for reporting analyses"
etic	available in advance, rather than determined during the analysis. Must be discovered, not predicted
emic	denoting subjective knowledge or experience
defining the population	deciding the entire group of individuals, items, or events from which a researcher wants to draw conclusions or make generalizations about
unitizing reliability	the consistency of observers in identifying boundaries and segments within a data continuum (like a transcript or recording) for analysis
cross-sectional vs. longitudinal populations	a cross-sectional population is studied at a single point in time to provide a snapshot of the population's characteristics at that moment, while a longitudinal population involves tracking the same individuals or group over an extended period to observe
availability-based vs. exposure-based populations	the key distinction is in the criteria used for inclusion: availability-based populations are who is accessible, while exposure-based populations are who has experienced a specific risk factor
archives	a collection of messages, usually well indexed
sampling	process of selecting a subset of cases for study from the larger population
probability sampling	for generalizable to some population of messages
"simple random sampling	1. Pulling cases ""out of a hat" 2. Sampling frame numbered, using a table w/ random table and using random number generator
systemic sampling	selecting every "xth" case, either from sampling frame list or in some flow of occurrence over time
cluster sampling	includes any probability sampling in which a group or set of messages is sampled together, usually for logistic reasons
stratified sampling	this type of probability sampling is where the sampling frame is segmented according to categories on some variables of prime interest to the researcher
multistage sampling	any probability sampling technique in which 2 or more sampling steps are used
nonprobability sampling	those generated by non random techniques, are undesirable and should be used only when no other option exist
convenience sampling	relies on selection of readily available cases
purposive or judgment sampling	researcher making a decision as to what cases he or she deems appropriate
quota sampling	like a nonprobability stratified sample, but key variable categories are identified and then a certain number of cases from each category are included in the study.
sample size	the number of participants or data points selected from a larger population to represent it statistically in a study
universal variables	characteristics that differentiate all stimuli
defining the medium	trying to itemize the characteristics of something that makes it that thing
null hypothesis	logical opposite of a predicted research hypothesis, it is a hypothesis that includes a prediction of no relationship between or among variables
research hypothesis	formal testing of research hypothesis is a deductive process progressing from theory to derived hypothesis to the testing of that hypotheses, the results of which reflect on the original theory
conceptual definition	relating to or based on mental concepts
directional hypothesis	specifying the directionality of the relationships among variables
nondirectional hypothesis	posits differences but not the direction of the differences
measurement theory (or classical test theory)	an observed test score is a combination of a "true score" and "error score" (X = T + E), where the true score represents what the person actually knows or possesses, and the error score represents random fluctuations or inconsistencies in the measurement
random error vs. nonrandom error	Random errors are unpredictable, chance variations in measurements that occur for different reasons in each trial, affecting precision. Non-random (or systematic) errors are consistent, predictable biases that affect all measurements in the same way or pr
bias	a disproportionate weight in favor of or against an idea or thing, usually in a way that is inaccurate, closed-minded, prejudicial, or unfair
validity	the consistency of a measure with a criterion
reliability	the internal consistency of a measure
accuracy	extent to which a measuring procedure is free of bias (nonrandom error)
precision	the fineness of distinction made between categories or levels of a measure
replicability	involves repeating a study with different cases or in a different context, checking to see if similar results are obtained each time
face validity	involves researchers looking at the content of the measurement items and advancing an argument that on its face the measure seems to identify what it claims
criterion validity	consists of methods that assess a measure's worth by examining its relation to some outside criterion.
content validity	the degree to which an assessment, like a test, completely and accurately measures all aspects of the specific concept or domain it's intended to assess
construct validity	requires that a new measure be administered to subject along with at least 2 other measures
"convergent vs. discriminant	convergent validity demonstrates that a measure strongly correlates with other measures of the same or similar constructs discriminant validity shows that it does not correlate strongly with measures of different, theoretically unrelated constructs"
coder debriefing	a structured discussion, often in healthcare, that occurs after a patient's "code" (resuscitation event) or a simulation, where the team reviews what happened to identify successes, areas for improvement, and to enhance teamwork, patient safety, and overa
coding scheme	a defined system for assigning symbols or codes to data, enabling conversion into a specific format (like binary) for storage, transmission, or analysis
nominal, ordinal, interval, ratio variables	Nominal data are categories with no inherent order (e.g., eye color). Ordinal data have a natural order but inconsistent intervals (e.g., education levels). Interval data have order and equal intervals but no true zero (e.g., temperature). Ratio data poss
unit of the data collection	the individual entity (person, household, organization, etc.) from which data is gathered in a study
CATA	computer coding
dictionaries	a set of word, phrased, parts, of speech, or any other word-based indicators that is used as the basis for a search
human coding	the use of people as coders, with each using a standard codebook and coding form, to read, view, or otherwise decode the target content and record the objective and careful observations on preestablished variables
intercoder reliability	amount of agreement or correspondence on a measured variable among 2 or more coders or raters
intracoder reliability	assesses whether coders can agree on the delineation of units of data collection when that is part of the coding protocol
unitizing reliability	a measure of how consistently researchers can identify the boundaries and segments of content to be analyzed, known as "units"
agreement vs. covariation	agreement assumes the measurements are meant to be the same, while covariation does not
chance-corrected agreement	makes assumption that some portion of coders' agreement is due to chance and adjusts for that
"threats to reliability	1. Poorly executed coding scheme 2. Inadequate coder training 3. Coder fatigue and Coder drift 4. Presence of a rouge coder"
codebook	A written set of coding categories, definitions, and decision rules that guides coders so they classify content consistently.
coder training	Structured practice where coders are taught how to apply the codebook, practice on sample texts, and receive feedback until reliable coding is achieved.
blind coding	a process where researchers analyze data without knowing the identity of the data's source to reduce bias
index	A composite measure created by combining multiple indicators into a single score representing a broader concept.
reliability for manifest vs. latent content	Reliability is typically higher for manifest content (objective, surface-level features) and lower for latent content (interpretive, meaning-based features) because latent judgments require more subjective judgment.
manifest	Observable, surface-level content that is directly countable (e.g., word frequency, presence of a feature).
latent	Underlying meanings, themes, or interpretations that require coder judgment (e.g., tone, sentiment).
pilot reliabilities	done on a randomly selected subsample of the total sample message pool before the study begins in earnest. This pilot testing should not be analyzed as a part of the final dataset.
final reliabilities	done on another randomly selected subsample during the full data collection, to fairly represent the coders’ performance throughout the study. This final reliability assessment is the one reported in your write-up
percent agreement	A simple reliability measure showing the percentage of times coders made the same coding decision.
Holsti's method	Measures agreement between coders who may not code the exact same data. When coders analyze the same sections, it works like percent agreement, but it still fails to adjust for chance agreement.
Scott's pi	An intercoder reliability statistic that corrects percent agreement by accounting for agreement expected by chance.
Cohen's kappa	A widely used intercoder reliability measure that corrects for chance agreement and is appropriate for two coders.
Krippendorff's alpha	A flexible reliability statistic that handles multiple coders, missing data, and different levels of measurement; widely recommended in content analysis.
Gwet's AC1	An alternative chance-corrected reliability coefficient that avoids kappa’s paradox (low kappa despite high agreement).
Spearman's rho	A rank-order correlation used to assess association between ordinal-coded data or the consistency of rankings between coders.
Intraclass correlation coefficient (ICC)	A reliability measure used for interval or ratio data that evaluates consistency or absolute agreement across multiple coders
Pearson correlation coefficient (r)	A measure of linear association between two interval/ratio variables; sometimes used to compare coder scores but not ideal for categorical coding
Lin's concordance correlation coefficient (CCC)	A statistic assessing agreement by combining measures of precision (correlation) and accuracy (closeness to a 45-degree perfect agreement line).
attenuation	Reduction in the observed relationship between variables caused by measurement error or low reliability.
nominal level measurement	Categorizes data into discrete groups with no numerical order (e.g., gender, political party).
ordinal level measurement	Ranks data in ordered categories without equal spacing between them (e.g., Likert rankings, class year).
internal level measurement	Measures variables with equal spacing between values but no true zero (e.g., temperature in °F, Likert-type scales treated as interval).
ratio level measurement	Contains equal intervals and a meaningful zero point, allowing all mathematical operations (e.g., height, response time).
categorical data	Data placed into categories; includes nominal and ordinal levels.
continuous variables	Variables with values on an interval or ratio scale that can take on any number within a range.
falsification of a theory	The principle that theories must be testable and structured so evidence can show them to be false.
tentativeness of a theory	The idea that theories are always provisional and may be revised with new evidence.
reliability	Consistency or stability of a measurement instrument over time or across observers.
false precision	Reporting overly detailed numerical values that imply more accuracy than the measurement supports.
fallacy of misplaced precision	Assuming precise numerical information is meaningful when the underlying construct or data are not that precise.
reliability coefficient	A numerical estimate (0–1) expressing the consistency of a measurement.
test-retest reliability	Stability of a measure over time, assessed by correlating scores from two administrations.
alternate forms (parallel forms) reliability	Consistency determined by administering equivalent versions of the same measure to the same people.
split-half reliability	Internal consistency estimate obtained by correlating two halves of a test (e.g., odd vs. even items).
Cronbach's coefficient alpha	A measure of internal consistency showing how well a set of items measures a single construct.
validity	The degree to which an instrument measures what it claims to measure.
face validity	The extent to which a measure appears to assess the intended construct based on surface judgment.
expert jury validity	Validity established through evaluation by subject matter experts who judge whether items reflect the construct.
criterion validity	Whether a measure correlates with an external criterion known to indicate the construct.
concurrent validity	A type of criterion validity assessed by correlating a measure with a criterion at the same time.
predictive validity	Criterion validity demonstrated when a measure predicts future behaviors or outcomes.
Construct validity	The extent to which a measure represents the theoretical construct and relates to other variables as predicted.
Thurstone equal appearing interval scales	Scaling method where judges rate how favorable items are, creating items spaced equally along an attitude continuum.
Likert scales	Multi-item scales measuring attitudes using agreement-level responses (e.g., strongly disagree → strongly agree).
item to total reliability	Correlation of each item with the total scale score to assess whether items fit well with the overall construct.
intercoder reliability	Agreement among coders in content analysis, verifying consistency of content classification.
Guttman scalogram	A cumulative scale in which agreeing with a strong item implies agreement with all weaker items.
semantic differential-type scales	Scales using bipolar adjective pairs (e.g., good–bad) to measure attitudes on multiple evaluative dimensions.
empirical	Based on observation or data rather than theory alone.
experiment	Research design where the researcher manipulates variables to establish causal relationships.
survey	A method of systematically collecting self-report data from respondents using questionnaires or interviews.
questionnaires	Structured instruments used to gather responses in surveys.
validity	(Repeated term) Accuracy of a measure in capturing the construct.
reliability	(Repeated term) Consistency/stability of measurement.
check questions	Items inserted to detect inconsistent or inattentive responding.
polarity rotation	Reversing the direction of some scale items to disrupt response patterns or response sets.
response set	A habitual pattern of answering (e.g., always agreeing) regardless of item content.
anonymity	Condition in which the researcher cannot identify participants from their data.
confidentiality	Condition in which participant identities are known but protected from disclosure.
focus groups	Group interviews where participants discuss topics, allowing researchers to study shared meanings and interactions.
funnel questions	Interview structure that starts with broad questions and narrows to more specific ones.
structured interviews	Interviews using standardized, predetermined questions asked in a fixed order.
unstructured interviews	Open-ended, flexible interviews where questions evolve during conversation.
network analysis	Method for studying relationships among individuals or groups by mapping connections.
nodes	The individuals, groups, or units represented in a network.
semantic networks	Maps of how concepts or words are linked in meaning within communication content.
survey nonresponse	When selected participants do not complete the survey; reduces data quality and representativeness.	Ignoring emailed survey link.
response rates	% of people contacted who actually complete the survey.	100 contacted → 40 complete → 40% response rate.
refusal rates	% of people who explicitly decline participation.	“No thanks, I’m not interested.”
cooperation rates	% of people who engage after being reached (not including unreachable participants).	60 reachable, 45 agree → 75% cooperation rate.
survey	A research method for collecting standardized info from individuals via questionnaires/interviews.	Polls, customer satisfaction surveys, academic questionnaires.
Survey research	Systematic collection of data from a sample to generalize to a population.	Using Qualtrics to study political attitudes.
respondent	The individual who provides survey data.	Anyone who completes the questionnaire.
Push poll	Biased “survey” designed to persuade rather than measure.	“Would you support Candidate X if you knew they were corrupt?”
questionnaire	Structured set of questions used to collect data.	Online or paper form.
Direct questions	Ask about the topic straightforwardly or explicitly.	“Do you support raising tuition?”
Indirect questions	Ask about sensitive topics more subtly to reduce bias.	“How do you think most students feel about raising tuition?”
Questions vs statements	Questions ask directly; statements ask participants to rate agreement.	“I feel stressed in class” (statement).
Open-ended questions	Allow respondents to answer in their own words.	“Describe your experience with online learning.”
Close-ended questions	Provide set response options.	Multiple choice, checkboxes, Likert scales.
Double-barreled questions	Ask two things at once.	“How satisfied are you with your advisor and classes?”
Leading questions	Suggest a “correct” or preferred answer.	“Don’t you agree the new policy is unfair?”
bias	Systematic error that skews results (wording, sampling, response).	Leading questions, unbalanced scales.
Response symmetry and balance	Response options should reflect equal numbers of positive/negative choices.	Poor: “Very satisfied / Satisfied / Neutral / Very dissatisfied” (missing “dissatisfied”).
Contingency questions	Questions answered only by those who meet certain criteria.	“Do you smoke?” If yes → “How many per week?”
Matrix questions	Sets of questions with same response options arranged in a grid.	Rating 10 statements on a 1–7 scale.
response-sets	Patterned responding unrelated to content.	Clicking “Strongly agree” for every item.
check questions	Items inserted to verify attention or consistency.	“Select ‘Disagree’ for this statement.”
polarity rotation	Switching scale direction to reduce response-sets.	Alternating between positive → negative phrasing.
univariate	Describes one variable at a time.	Mean stress score, distribution of age.
central tendency	Statistical measures identifying the “center” of a distribution.	Mean, median, mode.
mean	Arithmetic average.	(5 + 6 + 7) / 3 = 6
median	Middle value when data is ordered.	Values: 2, 3, 5, 7, 8 → median = 5
mode	Most frequent value.	Scores: 1, 1, 2, 3 → mode = 1
range	Highest minus lowest value.	Max = 20, Min = 10 → Range = 10
variance	Average squared deviation from the mean.	Larger variance = more spread.
Standard deviation	Square root of variance; average distance from the mean.	Higher SD = greater variability.
shape	Form of data distribution (normal, skewed, bimodal).	Skewed right = long tail on right.
deviation square	Squared difference between each score and mean.	(X – mean)² for each participant.
outlier	Value far outside the typical range.	Most scores: 10–15; one score = 90.
histogram	A bar graph showing the frequency distribution of a variable.	Plotting how many students rate online lectures from 1–7.
skew	Asymmetry in a distribution.	Most people rate a speaker highly → tail is on the low side.
Positive skew	Tail extends to the right; many low scores, few high.	Most respondents report low conflict, but a few report very high conflict.
Negative skew	Tail extends to the left; many high scores, few low.	Most people rate relational satisfaction as high, with a few very low scorers.
bimodal	Two peaks (two common score clusters).	Responses to political messaging might cluster in two ideological groups.
Shapes of distribution	Refers to whether data are normal, skewed, bimodal, etc.	Understanding shape impacts which statistics you choose.
mesokurtic	Normal “medium peaked” distribution.	Typical distribution of communication anxiety scores.
leptokurtic	Very peaked with heavy tails.	Most students score similarly on a listening exam, but a few extreme scores.
platykurtic	Flat distribution with light tails.	Very diverse social media attitudes with little consensus.
Normal distribution	Symmetrical, bell-shaped distribution with predictable probabilities.	Communication competence scores in a large, random sample often resemble this.
Standard normal distribution	Normal distribution transformed to mean = 0, SD = 1.	Z-scores of emotional regulation scores.
Z scores	Standardized value showing how far a score is from the mean (in SD units).	A student’s listening score is +1.5 SD above the class mean.
Pearson product-moment correlation	Measures linear relationship between two variables (−1 to +1).	Relationship between message harshness and perceived hostility (r = .62).
"Linear vs. nonlinear	Linear = consistent direction.
Nonlinear = curved (“U-shaped,” “inverted U”).	Stress vs. communication performance may be nonlinear (too little or too much stress is bad)."
sampling	Selecting a subset of a population for study.
sample	People you actually study.
population	Group you want to generalize findings to (e.g., all college students).
statistic	Value computed from a sample (e.g., sample mean).
parameter	Corresponding true value in the population.
bias	Systematic error in sampling or measurement.
Sampling error	Only surveying students who attend class.
Forms of sampling	Difference between sample statistic and population parameter due to chance.
random	Everyone has an equal chance.
Simple random sampling	Full random selection (e.g., random number generator).
Stratified random sampling	Divide population into groups → sample proportionally.	Proportionate sampling of communication majors and non-majors.
Cluster sampling	Randomly select groups/clusters first.	Select certain classes, then survey all students within them.
systematic	Select every nth person.
Convenience sampling	Survey whoever is easily available.
Quota sampling	Fill specific demographic quotas, but not randomly.
Purposive sampling	Select respondents based on criteria (e.g., crisis communicators, influencers).
Snowball sampling	Participants recruit others.
Probability	Likelihood of an outcome occurring.
Confidence interval	Range of values likely to contain the population parameter.
Margin of error	Amount of sampling error tolerated.
Standard error	Standard deviation of the sampling distribution.
binomial	Distribution of two possible outcomes (e.g., “agree”/“disagree”).
Standard error of a binomial	SE formula for dichotomous variables (p vs. q).
Hypothesis testing	Process of determining whether observed results differ from chance.
Null hypothesis	States “no effect/no relationship.”
Type I error	False positive, Rejecting H0 when it’s true.
Alpha risk	Probability of Type I error (.05 is typical).
Type II error	False negative, failing to reject H0 when an effect actually exists
Beta risk	Probability of Type II error.
"John Stuart Mill's criteria for making causal interferences	1. Covariation (variables change together) 2. Temporal order (cause precedes effect) 3. Elimination of alternatives"
Spurious relationship	Two variables appear related but are caused by a third variable.
Necessary condition	Must be present for an effect to occur.
Sufficient condition	If present, guarantees the effect.
empirical	Based on observation and data.
experiment	Manipulate an IV → measure effect on DV.
confounding	An uncontrolled variable that explains results.
Internal validity	Confidence that IV caused DV.
control	Procedures used to reduce confounding.
Control approaches
Elimination and removal	remove confounds entirely.
Holding constant	keep variables the same (e.g., same message length).
matching	pair participants on characteristics.
blocking	treat a confound as a factor in the design.
randomization	random assignment; gold standard.
Statistical control	control via regression, ANCOVA.
Research designs
True experiment or randomized experiment	Randomized; highest internal validity.
quasi-experiment	No random assignment; moderate internal validity.
non-experiment	Correlational or observational.
Signal enhancing designs	Increase the strength of the treatment/manipulation.
Noise reducing designs	Reduce variability/error in DV.
Factorial design
factors	independent variables
levels	conditions of each factor
Main effects	effect of each IV independently
Interactive effects	IVs combine to affect DV
Description of experiments with two or more independent variables
2X2	2-by-2 design or 2 X 2 design = two IVs with two levels each
3X3	3-by-3 design or 3 X 3 design = two IVs with three levels each
2X2X2	2-by-2-by-2 design or 2 X 2 X 2 design = three IVs with two levels each
2X3	2-by-3 design or 2 X 3 design = two IVs, one IV with two levels and the other with three levels
Spector	Associated with measurement theory used in communication.
Measurement error	Difference between observed score and true score.
reliability	Consistency of measurement.
test-retest	Same measure over time → consistent results.
Equivalent forms	Two different versions yield similar scores.
validity	Accuracy of measurement (measures what it claims).
generalizability	Extent findings apply to larger populations.
Internal validity	Confidence the IV caused the DV (repeated from above).
External validity	Generalizable beyond the study.
Sources of invalidity	participants change due to measurement
Instrument reactivity	unreliable instruments
history	outside events
Unreliable instruments
Differential subject loss	attrition not equal
Bias in assignment of subjects	nonrandom groups
External invalidity
Reactivity of instrumentation	observer effects
Hawthorne effects	participants change because they know they’re being studied
Invalidity of instruments	poor measurements
Confounding characteristics of particular samples	unique groups
"Design notation	R = random assignment
X = treatment
O = observation"
One-group designs	O X O
"Pretest-posttest design	R O X O
R O O"
Interrupted time series
Correlational designs
Cross-sectional design
Two-group design
Multiple-group posttest design
Multiple-group pretest posttest design
Ex post facto design
Multiple-group time series design
Factorial design
Solomon four-group design
Experiment	A research design where the researcher manipulates an independent variable and observes its effect on a dependent variable while controlling extraneous factors.	Testing whether message framing (gain vs loss) changes compliance in health communication.
Confounding	When an outside variable varies along with the IV and influences the DV, making it unclear what caused the effect.	Participants in one message-condition happen to be older than in the other.
Control	Procedures used to eliminate or reduce alternative explanations for observed effects.	Holding message length constant across conditions.
Counterbalancing	Alternating the order of conditions to reduce order or fatigue effects.	Half the participants see message A then B; the other half see B then A.
Matching	Pairing participants with similar characteristics and assigning them to different groups.	Matching dyads based on age or communication skill.
Blocking	Grouping participants by a variable and analyzing that variable as a factor.	Blocking participants by gender when testing ad effectiveness.
Randomization	Assigning participants to conditions by chance.	Randomly assigning people to “supportive message” or “critical message” groups.
Experimental Invalidity	Any threat that weakens confidence in causal conclusions from an experiment.	History, maturation, demand characteristics, etc.
Internal Invalidity	When observed effects may be due to factors other than the IV.	Participants in one condition have more prior knowledge.
External Invalidity	When results fail to generalize beyond the study sample or context.	Lab experiment may not generalize to real online conversations.
Factors	Independent variables in an experiment.	Message type, channel, emotional tone.
Levels	The different categories or values of a factor.	“High fear,” “moderate fear,” “low fear.”
Main Effects	The isolated effect of one factor regardless of other factors.	Emotional tone affects message recall regardless of channel.
Interaction Effects	When the effect of one factor depends on the level of another factor.	Fear appeals increase compliance only when the source is credible.
Ordinal Interactions	The direction of the effect stays the same, but its strength changes across levels.	High fear works better than low fear across all source types, but especially with high credibility.
Disordinal Interactions	The direction of the effect reverses depending on the level of another factor.	High fear works best with experts but low fear works best with peers.
Manipulation Check	A test within the study to confirm that the IV was perceived as intended.	Asking participants if they thought the message was “high fear.”
Research	A systematic process of gathering evidence to answer communication questions.	Studying how texting affects relational closeness.
Reasons for Conducting Research	To describe, explain, predict, or control communication phenomena.	Describing media use patterns; predicting effects of messages.
Overgeneralization	Drawing conclusions from too little evidence.	Assuming all students dislike Zoom after talking to two classmates.
Selective Observations	Seeing only what confirms preexisting beliefs.	Noticing only rude comments online when studying civility.
Inaccurate Observations	Misinterpreting or incorrectly recalling communication events.	Misremembering who initiated conflict in a conversation.
Illogical Reasoning	Reaching conclusions based on faulty logic.	“I got ghosted once, so everyone ghosting means the medium is toxic.”
Basic Research	Conducted to build or test theory without practical goals.	Studying how interpersonal conflict unfolds turn-by-turn.
Applied Research	Aimed at solving real-world communication problems.	Testing which message increases vaccination intentions.
Communication	The process of creating and interpreting messages within social contexts.	Interpersonal talk, media messages, organizational emails.
Message	Any symbolic representation used to convey meaning.	A text, speech, emoji, or ad.
Verbal Cues	Spoken or written language elements.	Word choice, phrasing, grammar.
Nonverbal Cues	Meaning conveyed without words.	Facial expression, emojis, tone.
Qualitative Methods	Methods that examine communication in depth, focusing on meaning.	Interviewing people about conflict experiences.
Quantitative Methods	Methods that measure variables numerically and analyze patterns.	Surveys measuring communication competence.
Survey Methods	Collecting standardized self-report data from respondents.	Measuring attitudes toward political ads.
Descriptive or Observational Surveys	Measure what exists without manipulating variables.	Tracking social media usage.
Content Analysis	Systematic coding of communication messages.	Analyzing news coverage tone.
Opinion Surveys	Assess individuals’ attitudes or beliefs.	Polling viewers about a TV debate.
Experimental Methods	Manipulating variables to test causal effects.	Changing message style to examine persuasion.
Serendipity	Unexpected findings that open new research paths.	Discovering an unpredicted relational effect while testing something else.
Arguments	Logical claims supported with evidence.	Building a case for a hypothesis.
Research Standards of Excellence	Criteria such as clarity, rigor, ethics, and validity.	Designing unbiased surveys.
Ethical Challenges	Risks of harm or deception in communication studies.	Experiments involving emotional messages.
IRB	Institutional Review Board; evaluates ethical safety of studies.	Approving a study involving sensitive interviews.
Consent	Participants agree after understanding study procedures.	Signing a disclosure form before a survey.
Confidentiality	Researcher knows identity but protects it from disclosure.	Storing names separately from responses.
Anonymity	Researcher cannot link responses to identities.	Anonymous online survey.
Problems (Research Problems)	Questions or issues a study seeks to address.	“Do supportive messages reduce stress after conflict?”
Variable	A characteristic that can vary across individuals or messages.	Message length, emotional tone, credibility.
Constant	A characteristic that does not vary in a study.	Using only 30-second messages.
Hypothesis	A specific, testable prediction about relationships between variables.	“Higher fear appeals increase intentions.”
Working Hypothesis	Initial assumption guiding early study planning.	Expecting positive association before precise testing.
Nondirectional Hypothesis	Predicts a relationship but not its direction.	“Message tone will affect trust.”
Directional hypothesis	A prediction that specifies the expected direction of the relationship between variables.	“More social media use will lead to higher anxiety.”
Null hypothesis	A statement that there is no relationship or difference between variables.	“Social media use has no effect on anxiety.”
Independent variable	The variable that is manipulated or categorized to examine its effect.	Amount of time spent on Instagram.
Dependent variable	The outcome that is measured in response to the independent variable.	Participants’ anxiety scores.
Crucial experiments	Experiments designed to clearly support one theory or explanation over competing ones.	A test that determines whether attention or memory primarily drives message recall.
Criteria for sound problem statements	Clear, specific, researchable, significant, and grounded in existing literature.	A precise statement explaining what phenomenon needs to be studied and why.
Requirements of hypotheses	Must be testable, falsifiable, specific, and grounded in theory.	A hypothesis linking texting frequency to relationship satisfaction.
Science	A systematic method of gathering knowledge through observation, testing, and evidence.	Conducting controlled studies of communication behavior.
Theory	An organized set of concepts and explanations that describe, predict, or explain phenomena.	Social Information Processing Theory explaining online impression formation.
Pseudoscience	Claims that appear scientific but lack empirical testing, falsifiability, or rigor.	Horoscopes predicting communication compatibility.
Scientific method	A structured process: problem identification, hypothesis formation, testing, analysis, and conclusion.	Designing a study to test how tone affects message interpretation.
Abstract calculus	Reasoning with abstract concepts rather than concrete events.	Using constructs like “identity” or “self-esteem” to build communication models.
Constructs	Abstract concepts studied in research but not directly observable.	“Self-disclosure” or “media trust.”
Rules correspondence	How well theoretical statements match empirical measures.	Ensuring your “communication apprehension” scale reflects the concept accurately.
Definitions	Statements that clarify what a term means.	Defining “online harassment” before measuring it.
Conceptual definitions	Theoretical explanations of what a concept means.	“Anxiety is a negative emotional state marked by worry.”
Daily definitions	Informal, everyday meaning of a concept.	Defining “stress” as “feeling overwhelmed.”
Poetic definitions	Figurative, expressive explanations.	“Anxiety is a storm inside the mind.”
Scholarly definitions	Academic, precise, literature-based definitions.	“Stress is a psychophysiological response to perceived demands exceeding coping resources.”
Circularity	A faulty definition that uses the term itself in the definition.	“Influence is when someone influences another person.”
Operational definitions	How a concept is measured or manipulated in a study.	Measuring stress via a 7-item stress scale.
Forms of operational definitions	Measurements (scales), observations, manipulations, categorization.	Manipulating message tone (friendly vs. hostile).
Manipulated independent variables	Variables the researcher actively changes.	Sending participants either a supportive or critical message.
Measured/assigned variables	Variables observed or classified rather than manipulated.	Gender, personality scores, or political ideology.
Direct classification variables	Variables sorted into categories without manipulation.	Categorizing participants as “frequent” vs. “infrequent” texters.
Intersubjectivity	Agreement among multiple observers that a measurement or observation is consistent.	Coders rating a message as sarcastic with high agreement.
Repeatability	The ability to replicate findings using the same methods.	A study on message tone yielding similar results in another lab.
Standards for operational definitions	Must be reliable, valid, precise, and replicable.	A depression scale consistently measuring depression across samples.
Criticism of conceptual definitions	They can be vague, abstract, or difficult to test empirically.	“Motivation” can be defined so broadly that it’s hard to measure.
Inductive reasoning	Drawing general conclusions from specific observations.	Observing multiple hostile comments and proposing a theory of online aggression.
Deductive reasoning	Starting with theory and using it to generate predictions.	Using attachment theory to predict texting behaviors.
Requirements of theory	Should be testable, falsifiable, parsimonious, logically consistent, and useful.	A communication theory that offers clear, testable predictions.
Falsification	The ability for a theory to be proven wrong through testing.	A theory predicting “all” messages lead to the same interpretation can be disproven.
Tentativeness	Scientific knowledge is provisional and may change with new evidence.	Modifying a theory when new data contradicts older findings.
Functions of theory	Describe, explain, predict, and control phenomena.	A messaging theory helping identify what communication strategies reduce conflict.
Description	Identifying and detailing communication behaviors.	Documenting how often partners text daily.
Explanation	Clarifying why a behavior occurs.	Explaining why people use emojis to soften tone.
Prediction	Forecasting outcomes based on theory.	Predicting that supportive messages increase relationship satisfaction.
Control	Using theories to shape or guide outcomes.	Designing interventions to reduce misinformation sharing.
Argument	A set of statements used to support a claim through logic and evidence.	Writing a research paper defending a communication theory.
Abstract	A brief summary of a research article’s purpose, methods, results, and conclusions.	A 150-word paragraph summarizing your study on social media stress.
Plagiarism	Using another’s work without proper credit.	Copying definitions from a textbook without citation.
Exhaustive literature reviews	Reviews that cover all relevant research on a topic comprehensively.	A dissertation chapter summarizing every major study on media anxiety.
Exemplary literature reviews	Reviews that are thorough, analytical, well-structured, and evaluative.	A review that connects themes, critiques methods, and synthesizes theories.
Explication	Clarifying and refining concepts by breaking them into precise components.	Explaining the subdimensions of “self-disclosure” in online communication.
Research prospectus	A brief plan that explains what you want to study, why it matters, and how you will study it.	A student outlines a project examining how relational uncertainty affects texting conflicts.
Evidence	Information that supports a claim or argument in communication research.	Using survey data to show that supportive messages reduce stress.
Ecological fallacy	Incorrectly assuming that group-level patterns apply to every individual.	Seeing that a class does well on average and assuming every student understands the material.
Delphi fallacy	Believing an idea is true because experts agree, even when data is weak.	A panel of professionals predicts a media trend, but no empirical support exists.
Jeanne Dixon fallacy	Remembering correct predictions but ignoring incorrect ones.	A commentator remembers when they “predicted” a political debate outcome but ignores the times they were wrong.
Patchwork quilt fallacy	Treating unrelated cases as if they form a meaningful pattern.	Linking a few isolated viral TikTok posts to claim a communication trend exists.
Factual evidence	Verifiable information based on observation or measurement.	Reporting the percentage of viewers who watched a news broadcast.
Primary sources	Original research or first-hand data.	A published experiment on message framing.
Secondary sources	Summaries or reviews of primary research.	A textbook chapter explaining uncertainty reduction theory.
Reports	Descriptive summaries of events or conditions without interpretation.	A report showing the number of complaints sent to a customer service center.
Statistics	Numerical summaries used to describe communication patterns.	Reporting the average length of emails in a workplace study.
Parameters	Numbers that describe an entire population.	The true average time all employees spend in meetings.
Opinions	Personal judgments or beliefs not based on empirical evidence.	Saying “people prefer short texts” without data.
Expert opinions	Judgments from qualified professionals based on their knowledge.	A PR specialist describing best practices for crisis messaging.
Inductive reasoning	Moving from specific observations to general conclusions.	Coding several conflict conversations and forming a theory about escalation patterns.
Deductive reasoning	Testing a general idea with specific observations.	Starting with politeness theory and checking how often speakers soften criticism.
Causal argument	Claims that one communication factor produces change in another.	Arguing that supportive messages lower stress after exposure to hostile comments.
Syllogism	A structured argument with a major premise, minor premise, and conclusion.	Major: Supportive messages reduce stress. Minor: This message is supportive. Conclusion: It reduces stress.
Categorical syllogism	A syllogism using categories or groups.	“All deceptive messages omit information; this message omits information; therefore it is deceptive.”
Disjunctive syllogism	Uses “either/or” statements.	“Either the message is sarcastic or sincere; it is not sarcastic; so it is sincere.”
Conditional syllogism	Uses “if/then” reasoning.	“If a message offers empathy, the receiver feels supported; this message offers empathy; the receiver feels supported.”
Necessary condition	A condition that must be present for something to occur.	For persuasion, attention is necessary. No attention = no persuasion.
Sufficient condition	A condition that guarantees an outcome.	A credible source may be sufficient to produce trust in some settings.
Measurement	Assigning numbers to communication behaviors or attitudes.	Rating message clarity on a 1–7 scale.
Measurement error	Any difference between the true value and the measured value.	A participant misreads a survey item and picks the wrong response.
Attenuation	Weakening of relationships due to measurement error.	Poorly worded items reduce the correlation between media use and anxiety.
Signal-to-noise ratio	Amount of meaningful information compared to error.	Clear, well-designed items increase signal; unclear ones add noise.
Nominal level measurement	Categories with no numerical meaning.	Coding messages as supportive, neutral, or hostile.
Ordinal level measurement	Ranking without equal spacing.	Ordering messages from least to most polite.
Interval level measurement	Equal spacing but no true zero.	A 1–7 Likert scale on topic involvement.
Ratio level measurement	Equal spacing with a true zero.	Counting seconds of silence in a conversation.
Continuous data	Data with many possible values.	Measuring reaction time after seeing a message.
Categorical data	Data in distinct categories.	Coding a post as “informational” or “emotional.”
Reliability	Consistency of a measure.	A politeness scale produces similar results when retesting.
Reliability coefficient	Numerical estimate of measurement consistency.	Cronbach’s alpha = .88 shows good internal consistency.
Test-retest reliability	Stability of scores over time.	Participants complete a media literacy scale two weeks apart.
Alternate forms reliability	Consistency between two versions of a measure.	Two versions of a social presence scale produce similar scores.
Split-half reliability	Consistency between halves of a measure.	Odd items correlate with even items on a deception scale.
Intercoder reliability	Agreement among coders analyzing communication content.	Two coders classify political ads and match 90% of the time.
Item-to-total reliability	Correlation between each item and the full scale.	One weak item on an empathy scale shows a low item-total correlation.
Validity	How well a measure reflects what it claims to measure.	A credibility scale actually measures source credibility, not liking.
Face validity	Whether items appear to measure the construct.	“This scale looks like it measures conflict avoidance.”
Expert jury validity	Experts judge whether items fit the construct.	Scholars review a new parasocial interaction scale.
Criterion validity	Relationship between the measure and an external standard.	A loneliness measure correlates with time spent alone.
Concurrent validity	Correlates with another measure at the same time.	Two immediacy scales correlate strongly.
Predictive validity	Predicts future outcomes.	Conflict style predicts later relational satisfaction.
Construct validity	Overall evidence that a measure represents the concept.	Politeness, face, and respect items work together coherently.
Semantic differential	Bipolar adjective scale.	Rating a message as warm–cold or friendly–unfriendly.
Likert scale	Agreement-based scale.	“I feel supported by this message” (1–7).
Guttman scalogram	Items arranged in increasing intensity.	Agreeing to stronger items implies agreement with weaker ones.
Thurstone scale	Items selected based on expert-judged intensity.	Experts rate hostility levels to build a hostility scale.
Negative case analysis	Looking for data that contradicts a pattern.	A study on supportive messages includes examples that do not help recipients.
Population	The full group a researcher cares about.	All college students who use Instagram for news.
Sampling	Selecting a smaller group from the population.	Recruiting 300 students from several campuses.
Data	Information collected during research.	Survey answers about online conflict.
Statistic	A number describing a sample.	The sample mean level of message clarity.
Parameter	A number describing a population.	The true average clarity rating among all students.
Bias	Systematic error in sampling or measurement.	Only surveying students who volunteer results in selection bias.
Representative sample	Sample that reflects the population’s characteristics.	A sample matched to the population on age, gender, and media use.
Sampling error	Random differences between the sample and population.	The sample reports 30% Instagram news use, but the real rate is 27%.
Sampling error	The difference between sample results and the true population values due to chance.	A sample of students overestimates average texting frequency compared to the whole campus.
Standard deviation	A measure of how spread out scores are around the mean.	Variation in how many TikToks participants watch daily.
Probability sampling	Sampling methods where each member of the population has a known chance of being selected.	Using random selection to choose participants for a media effects study.
Random sampling	Every member of a population has an equal chance of selection.	Randomly selecting phone numbers to study call habits.
Stratified random sampling	Population is divided into strata, and participants are randomly selected from each group.	Sampling equal numbers of first-year and senior students to compare online disclosure.
Cluster sampling	Groups (clusters) are randomly selected instead of individuals.	Randomly choosing entire classrooms to study social media habits.
Systematic or periodic sampling	Selecting every nth unit from a list.	Surveying every 10th student entering the student union about their preferred messaging apps.
Accidental or convenience sampling	Selecting whoever is easiest to reach.	Surveying students in the library about texting anxiety.
Quota sampling	Nonrandom selection but with quotas for key categories.	Ensuring half the sample uses Instagram and half uses TikTok.
Purposive or known group sampling	Selecting participants based on specific characteristics.	Recruiting people who frequently engage in online debates.
Snowball sampling	Participants recruit additional participants, useful for hard-to-reach groups.	Asking cyberbullying victims to refer others for a study.
Descriptive statistics	Statistics that summarize data.	Reporting the average number of daily texts sent.
Inferential statistics	Methods used to make conclusions about populations based on samples.	Testing whether message tone affects perceived credibility.
Central tendency	Measures describing the typical score in a distribution.	Reporting the typical number of emojis used per message.
Arithmetic mean	The average score of all values.	Average number of YouTube videos watched per day.
Unbiased estimator	A statistic whose expected value equals the population parameter.	Using sample variance to estimate true population variance.
Median	The middle score in an ordered list.	The median number of hours spent on social media per day.
Mode	The most frequently occurring value.	The most common messaging app used by participants.
Bimodal	A distribution with two modes.	A study finds participants mostly prefer either texting or Snapchat, with few choosing other platforms.
Measures of variability or dispersion	Statistics that show how spread out data are.	Examining variation in conflict frequency in online relationships.
Range	Difference between highest and lowest scores.	Range of “hours on TikTok” in the sample: 0–8 hours.
Variance	Average squared deviation from the mean.	Variance in participants’ communication apprehension scores.
Standard deviation	The square root of variance; shows typical spread.	Daily screen time varies with an SD of 1.8 hours.
Skewness	A measure of asymmetry in a distribution.	Social media use positively skewed because a few people use it excessively.
Kurtosis	Describes how peaked or flat a distribution is.	Message length distributions that are highly peaked or flat.
Platykurtic	Flat distribution; light tails.	Wide variation in how often people check notifications.
Mesokurtic	Normal distribution; moderate tails.	Typical distribution of daily texts sent.
Leptokurtic	Peaked distribution; heavy tails.	Most participants send about the same number of messages, but a few send extremely many.
Standard normal curve	A normal distribution with mean 0 and SD 1.	Z-score distributions of media anxiety scores.
Probability distribution	Shows the likelihood of each possible value of a variable.	Probability of different levels of self-disclosure in messages.
Data distributions	Patterns showing how data values are arranged.	Histogram of how often people post on Instagram weekly.
z scores	Standardized scores representing how many SDs a value is from the mean.	A participant’s TikTok usage is +2 SDs above the mean.
Correlation	Statistical measure of the relationship between two variables.	More texting is correlated with closer friendships.
Direct relationship	Variables increase or decrease together.	As supportive messages increase, satisfaction increases.
Inverse relationship	One variable increases as the other decreases.	More hostile comments lead to less willingness to respond.
Positive relationship	Both variables move in the same direction.	More memes shared → higher relational closeness.
Negative relationship	Variables move in opposite directions.	More screen time → lower sleep quality.
Curvilinear relationship	Relationship changes direction at different levels.	Moderate self-disclosure increases liking, but excessive disclosure decreases it.
Guidelines to interpret correlations	r ≈ .10 small, .30 medium, .50 large (context-dependent).	r = .45 between texting frequency and closeness = strong in communication research.
Coefficient of determination (r²)	Proportion of variance explained by the correlation.	r = .50 → r² = .25: 25% of closeness is explained by texting frequency.
Pearson product moment correlation	Correlation for interval/ratio variables.	Relationship between hours online and stress level.
Spearman rank order correlation	Correlation for ranked/ordinal data.	Rank-ordering participants by communication apprehension.
Outlier	A value far from others in the dataset.	One respondent reports sending 10,000 texts per day.
Probability	The likelihood an event occurs.	Probability a random person prefers TikTok over Instagram.
Probability distribution	How probabilities are assigned to all possible outcomes.	Distribution of likelihood of different message lengths.
Research hypothesis	A prediction about relationships or differences between variables.	“Supportive messages will increase perceived trust.”
Null hypothesis	States that no effect or relationship exists.	“Supportive messages have no effect on trust.”
Critical region	Area of the distribution where the null is rejected.	The most extreme 5% of message sentiment scores.
Significant difference or relationship	A result unlikely due to chance (p < .05).	Friendly tone significantly boosts message credibility.
Critical value	Cutoff value determining whether results fall in the critical region.	t must exceed ±2.00 to be significant.
Standard error of the mean	The SD of sample means; measure of sampling precision.	SEM shows how precisely we estimate average social media use.
Type I error	Rejecting the null hypothesis when it is actually true.	Concluding that emoji use changes trust when it actually does not.
Type II error	Failing to detect an effect that actually exists.	A study concludes a supportive message doesn’t reduce stress when it actually does.
Alpha risk	The probability of making a Type I error.	Setting α = .05 means accepting a 5% chance of falsely claiming a message strategy works.
Beta risk	The probability of making a Type II error.	High β means a study might miss the effect of media exposure on fear.
Power	The ability to detect an effect when it exists (1 – β).	A large sample gives enough power to detect small differences in message clarity.
Test statistic	A number used to decide whether to reject the null hypothesis.	A t-test value comparing message warmth ratings between groups.
One-tailed tests	Tests predicting the direction of an effect.	Hypothesis: a polite message will produce more trust than a blunt message.
Two-tailed tests	Tests predicting any difference, not a direction.	Hypothesis: message style influences trust, but direction is not predicted.
Parametric tests	Statistical tests that assume normal distribution and equal variances.	Using ANOVA to compare persuasion across message conditions.
Assumptions underlying parametric tests	Normal distribution, equal variances, interval/ratio data.	Checking if message valence scores are normally distributed before a t-test.
Ceiling or floor effects	Scores cluster at the high or low end, limiting variability.	A politeness scale is too easy, and everyone scores near the maximum.
Subjects by treatment interactions	When participants respond differently to the same treatment.	Some people benefit from humor in messages; others find it annoying.
Correlation	A relationship between two variables.	Time spent online correlates with perceived social support.
Effect size	The strength of a relationship or difference.	Supportive messages have a medium effect size on reducing stress.
Analysis of variance (ANOVA)	Tests whether means differ across three or more groups.	Comparing persuasion from emotional, logical, and neutral messages.
Main effects	The effect of one independent variable, ignoring the other.	Emotional messages increase persuasion across all audience types.
Interaction effects	When the effect of one variable depends on another.	Humor increases persuasion only for low-involvement audiences.
Nonparametric tests	Tests that do not assume normality.	Using chi-square to analyze categories of conflict strategies.
Multicollinearity	When predictors are strongly correlated, making analysis unstable.	Source credibility and message quality overlap too much in a regression.
Chi-square test	A test comparing expected vs. observed frequencies.	Examining whether men and women differ in apology types.
Factor analysis	Identifies clusters of related items forming underlying dimensions.	A scale reveals two factors: emotional support and informational support.
Nomothetic research	Seeks general patterns across many cases.	Studying media effects using large-sample surveys.
Idiographic research	Seeks detailed understanding of one case.	An in-depth interview study about one online community.
Text	Any communication artifact that can be analyzed.	A speech, meme, ad, or political tweet.
Rhetorical criticism	Analyzing messages to understand how they persuade or shape meaning.	Interpreting a campaign speech using rhetorical theory.
Rhetoric	Strategic use of symbols to influence an audience.	A climate activist’s speech aimed at motivating action.
Impressionistic criticism	Personal, subjective evaluations of a text.	Saying a speech “felt sincere and inspiring” without systematic analysis.
Straw man fallacy	Misrepresenting an argument to attack it more easily.	Saying “opponents want no regulation at all” when they only want moderate change.
Neo-Aristotelian criticism	Evaluating rhetoric through the five classical canons.	Analyzing a TED talk’s invention, arrangement, and delivery.
Invention	The ideas and arguments a speaker uses.	Choosing evidence to support a claim about media bias.
Ethos	Credibility and trustworthiness.	A doctor discussing vaccine information.
Pathos	Emotional appeals.	A charity ad showing families impacted by disaster.
Logos	Logical reasoning.	Presenting statistics on texting-and-driving accidents.
Arrangement	How ideas are organized.	Opening with a story, then evidence, then a call to action.
Style	Language choices and rhetorical devices.	Using metaphors to frame climate change as a “ticking clock.”
Delivery	The physical presentation of the message.	A speaker’s pacing, eye contact, and vocal tone.
Memory	Mastery of the material and ability to present fluidly.	A speaker recalls their speech without notes.
Identification	Building a sense of shared identity with an audience.	A speaker highlights common values: “We all want safer communities.”
Burke’s dramatistic pentad	A tool analyzing motives using five elements.	Analyzing a crisis response using act, scene, agent, agency, purpose.
Act	What happened.	“The company delayed reporting the data breach.”
Scene	The context in which it happened.	“During a period of rising cybersecurity threats…”
Agent	Who performed the act.	“Company leadership concealed the breach.”
Agency	The means used.	“They used internal memos instead of public statements.”
Purpose	Why the act occurred.	“To prevent a loss of public trust.”
Fisher’s narrative paradigm	Communication is storytelling; good stories persuade.	A politician’s story resonates because it feels authentic.
Narrative fidelity	Whether a story matches lived experience and values.	A testimony about hardship feels true to audience experience.
Utterance	A single spoken or written unit of communication.	One sentence in a political debate.
Discourse analysis	Studying language use in context.	Examining how people soften criticism in email.
Fieldwork	Collecting data in natural settings.	Observing communication in a community center.
Participant observation	Researcher interacts while observing.	Joining a fan community to study their communication norms.
Naturalistic studies	Studying communication in real-life settings.	Observing small group meetings instead of using a lab.
Full participant observation	Researcher fully participates and conceals their role.	Joining a workplace chat channel without disclosing researcher status.
Participant-as-observer research	Researcher participates and reveals their role.	Facilitating workshops while noting interaction patterns.
Etic approach	Studying culture from an outsider’s viewpoint.	Analyzing communication norms in a community you are not part of.
Emic approach	Studying culture from an insider’s viewpoint.	Using community members’ perspectives to explain group conflict.
Accretions	Physical traces left behind.	Old flyers showing how a student group communicated over time.
Erosions	Wear-and-tear traces showing use patterns.	Worn-out seats in areas where people gather to talk.
Grounded theory	Building theory from data instead of starting with hypotheses.	Coding interviews to develop a model of online support exchange.
Ethnography	Detailed study of communication in a culture or group.	A year-long study of communication in activist organizations.
Time-budgeting studies	Tracking how people spend their time.	Logging how often teens check social media each hour.
Community studies	Examining communication practices in a community.	Studying how a neighborhood communicates during crises.
Case studies	In-depth analyses of a single case.	Analyzing one viral hashtag movement.
Negative case study	Examining cases that contradict the main pattern.	A support group that thrives without typical leadership structure.
Coding	Categorizing qualitative or quantitative data.	Coding text messages for emotional tone.
Open coding	First step of coding; identifying initial concepts.	Labeling segments as “support,” “conflict,” or “advice.”
Focus group	Guided group discussion used to gather attitudes.	A group discusses reactions to a health campaign message.
Content analysis	Systematic coding of texts to identify patterns.	Analyzing news articles for frames used in immigration coverage.
Coding units	The elements being coded.	Each tweet is one coding unit.
Latent content	The underlying meaning.	Coding sarcasm in online posts.
Manifest content	The surface-level content.	Counting mentions of specific keywords.
Reliability	Consistency of measurement.	Coders consistently label messages as supportive.
Validity	Accuracy of the measure.	A scale truly measures communication apprehension.
Empirical	Based on observation and evidence.	Collecting data on how students use discussion boards.
Experiment	Manipulating variables to test cause and effect.	Testing whether supportive messages reduce stress after conflict.
Survey	Asking participants questions to measure attitudes or behaviors.	A survey on media use and loneliness.
Questionnaires	Structured sets of survey questions.	Asking respondents to rate message clarity on a scale.
Check questions	Items used to detect inattention or inconsistency.	“Select ‘strongly agree’ for this item.”
Reliability	The consistency or stability of a measurement across time, items, or observers.	A communication apprehension scale gives similar results when taken two weeks apart.
Check questions	Items inserted to verify respondent attention or honesty.	“Select ‘strongly agree’ for this item” to detect careless responding in a media survey.
Polarity rotation	Reversing the direction of response options to prevent patterned responding.	Alternating between positively and negatively worded items in a trust scale.
Response set	A habitual way of answering regardless of content.	Participants repeatedly choosing “agree” on all items in a social media attitudes survey.
Anonymity	No identifying information is collected; responses cannot be linked to participants.	Online survey of cyberbullying experiences with no email or name recorded.
Confidentiality	Researcher knows identities but does not disclose them.	Interviewing influencers about harassment and keeping their identities protected.
Funnel questions	Questions that start broad and become increasingly specific.	Interviewing about “general online habits” before asking about “TikTok body image effects.”
Structured interviews	Interviews using fixed, standardized questions.	Asking each participant identical questions about message credibility.
Unstructured interviews	Open-ended, conversational interviews guided by participant responses.	Letting participants discuss their experiences with online conflict in their own way.
Experiment	A controlled study where variables are manipulated to establish causality.	Testing how different message tones (friendly vs. hostile) affect perceived warmth.
Confounding	When another variable unintentionally influences the results.	Message length differs between tone conditions, making results unclear.
Control	Holding variables constant to isolate the independent variable’s effect.	Keeping message length identical across tone conditions.
Counterbalancing	Varying the order of conditions to control order effects.	Seeing a supportive message first vs. hostile message first.
Matching	Equating participants on key characteristics across groups.	Matching groups based on prior social media use before testing message effects.
Blocking	Creating subgroups based on a characteristic and randomly assigning within each.	Blocking by gender before assigning participants to message tone conditions.
Randomization	Assigning participants to conditions by chance.	Randomly assigning participants to emoji vs. no-emoji messaging conditions.
Experimental invalidity	Any factor that threatens the accuracy of experimental conclusions.	Participants guessing the study purpose and altering responses.
Internal invalidity	The extent to which results are due to the IV and not confounds.	If message length differs by condition, internal validity is reduced.
External invalidity	The extent to which results generalize beyond the study.	Lab results on message tone may not generalize to real texting behavior.
One-shot case study	Treatment → measurement with no comparison group or pretest.	Show a persuasive ad, then measure attitude.
One-group pretest-posttest	Pretest → treatment → posttest but no control group.	Measure attitudes → show a political ad → measure attitudes again.
Static-group comparisons	Two groups: one receives treatment, the other does not; no randomization.	One class sees an anti-cyberbullying message; another class does not.
Pretest-posttest control group design	Random assignment with both pretest and posttest.	Test trust before and after seeing supportive messages, with a control group.
Solomon four-group design	Combines pretest/posttest and posttest-only groups to control testing effects.	Four groups receive combinations of pretests and supportive/neutral messages.
Posttest-only control group design	Random assignment, treatment vs. control, posttest only.	Participants get a supportive message or neutral message, then rate trust.
Factorial designs	Experiments with two or more independent variables.	Testing message tone (friendly/hostile) × emoji use (yes/no).
Factors	Independent variables in a factorial design.	Tone and emoji use.
Levels	The categories or conditions within each factor.	Tone has two levels: friendly and hostile.
Main effects	The independent effect of each factor on the DV.	Friendly messages increase trust regardless of emoji use.
Interaction effects	Effects that occur when one factor depends on another factor’s level.	Emojis increase trust only when tone is friendly.
Ordinal interactions	Interaction where effects have the same direction but different magnitude.	Friendly messages always increase trust, but especially when emojis are used.
Disordinal interactions	Interaction where effects reverse direction.	Emojis increase trust in friendly messages but decrease trust in hostile messages.
Manipulation check	A test to verify the independent variable worked as intended.	Asking participants whether the message seemed “friendly” or “hostile.”
Pilot studies	Small, preliminary studies used to refine procedures.	Testing whether participants correctly interpret the message stimuli.
Scott’s pi	A measure of intercoder reliability correcting for chance.	Two coders categorize comments as supportive/critical with Scott’s pi = .85.
Krippendorff’s alpha	A reliability statistic usable with any number of coders and data types.	Alpha = .90 for coding sarcasm in tweets.
Cohen’s kappa	Intercoder reliability for two coders correcting for chance.	Two coders rating message tone reach κ = .82.
Meta-analysis	A statistical technique that synthesizes results across many studies.	Combining 40 studies on cyberbullying and depression.
Primary analysis	Analyzing data that you directly collected.	Running statistics on your own experiment on emoji use.
Secondary analysis	Reanalyzing existing data collected by others.	Using Pew Research datasets to examine TikTok usage patterns.
Narrative reviews	Qualitative summaries of past research without statistical synthesis.	Writing a literature review on online self-disclosure.
Assumptions of meta-analysis	Requires independence of studies, comparable measures, and adequate reporting.	Each study on message tone must treat participants as separate samples.
Independence	Each dataset must be statistically separate from others.	Two studies using the same participant pool violate independence.
Effect size	A quantitative measure of the strength of a relationship. r and d	Reporting that message warmth has a medium effect on trust.
r	Correlation-based effect size.	r = .40 between texting frequency and closeness.
d	Difference-based effect size (mean differences).	d = .60 for supportive vs. neutral message effects.
Publication bias	Tendency for studies with significant results to be published more often.	Journals publish “message tone affects credibility,” but not null results.
File drawer effect	Non-significant studies remain unpublished and “in the drawer.”	Researchers do not publish studies finding no link between screen time and stress.
Fail-safe number	Number of hypothetical null studies needed to eliminate a significant meta-analysis result.	A meta-analysis reports that it would take 150 null studies to negate effects of cyberbullying on anxiety.
Advantages of meta-analysis	Increases statistical power, summarizes large bodies of research, identifies patterns.	Clarifying the overall effect of online support on well-being across dozens of studies.
Disadvantages of meta-analysis	Vulnerable to publication bias, varying study quality, and non-equivalent measures.	A meta-analysis of “social media use” suffers because studies define the construct differently.

Created by: user-2006078

"Know" box contains:
Time elapsed:
Retries: