Student Center | Instructor Center | Information Center | Home
Communication Research
Student Center
Internet Resources
Glossary
Flashcards
Research Examples


Feedback
Help Center



Keyton: Communication Research

Glossary

alpha level  The significance or probability level set by the researcher for each statistical test prior to conducting the study; in communication research, most often set at .05; represented by the symbol p.
analysis of variance  Statistical test compares the influence of two or more groups of one or more nominal independent variables on a continuous level dependent variable; represented by the symbol F; also referred to as ANOVA.
analytic induction  Reasoning process used with qualitative research methods; moves from the specific to the general; researchers discover and develop theories as they emerge from the data.
anonymity  Protection of names and other pieces of information that can identify participants; researchers do not ask participants to reveal information that would aid the researcher in identifying participants' individual data.
ANOVA  Statistical test that compares the influence of two or more groups of one or more nominal independent variables on a continuous level dependent variable; represented by the symbol F; also referred to as analysis of variance.
antecedent variable  Variable manipulated by the researcher; presumably, this manipulation, or variation, is the cause of change in other variables; also referred to as independent variable, experimental variable, treatment variable, and causal variable.
attrition  Threat to the internal validity of research when participants can no longer be observed or used for data collection because they have dropped out of the research project; also known as mortality.
authorial voice  The person(s), researcher or the participants, who tells the story in a qualitative research report.
behavior coding  Type of questionnaire pretesting for face-to-face surveys; a third person monitors the interaction between the interviewer and respondent to look for problems in the questionnaire or its administration.
beneficence  Protection of the well-being of participants; the researcher must meet the obligation to maximize possible benefits while minimizing possible harms.
beta coefficients  Unit of standardized scores in regression; indicates the difference in a dependent variable associated with an independent variable; also known as beta weights.
beta weights  Unit of standardized scores in regression; indicates the difference in a dependent variable associated with an independent variable; also known as beta coefficients.
between-groups variance  Variation of scores between categories or groupings of independent variable sufficient to distinguish themselves from one another.
between-subjects design  Design feature of ANOVA in which each participant is measured at only one level, group, or category, or under only one condition; designed to examine differences across individuals in the study.
biased  Favoring one attribute or characteristic more than another.
Boolean operators  Words used in database searches to connect or combine words and concepts; most common are AND, OR, and NOT. AND retrieves records or citations that include all of the combined search terms (for example, children AND television). OR retrieves all records with at least one of the descriptions (for example, children OR kids); NOT excludes the specified term from the search (for example, children NOT teenagers).
categorical data  Nominal data; form of discrete data; describes the presence or absence of some characteristic or attribute.
categorizing reliability  Degree to which multiple coders make similar distinctions among the data to be coded in assigning data to categories; used in content analysis.
causal variable  Variable manipulated by the researcher; presumably, this manipulation, or variation, is the cause of change in other variables; also referred to as antecedent variable, experimental variable, treatment variable, and independent variable.
census  Situation where every element of a population is included in a research project.
central tendency  Term applied to any of several measures that summarize a distribution of scores; mean, median, and mode are common measures of central tendency; this one number acts as a summary of all the scores on one variable.
chi-square  Represented by the symbol χ2; statistical test used to determine if differences among nominal, or categorical, level data are statistically significant; examines the observed frequencies in comparison to the expected frequencies to determine if the categorical differences that occurred are the same as would occur by chance.
classical experiment  Research design in which participants are randomly selected and the researcher controls the treatment or the manipulation of the independent variable by randomly assigning participants to treatment or control groups.
closed question  Question form in which respondents are asked a question (or given a statement) and then given a set of responses to select from.
cluster sampling  Form of random, or probability, sampling used when researchers do not have access to a complete list of population members; a two-stage or multistage process; in the first stage, groups, or clusters, of the population are selected; then simple random sampling is used in each cluster to select the research sample.
coefficient of determination  The percentage of variance two variables have in common; represented by the symbol r2; found by simply squaring the r value.
cognitive approach to pretesting  Type of questionnaire pretesting in which researcher asks questions about the stimulus question to eliminate alternative meanings.
Cohen's kappa  Measure of interrater reliability for categorical data; ranges from perfect agreement (1.00) to agreement that is no better than would be expected by chance (0).
complete observer  Form of participant observation; the researcher enacts only the observer role; the researcher is hidden completely from the interactants.
complete participant  Form of participant observation; researcher is a fully functioning member of the scene; the research role is disguised through the role as participant.
concept  Abstract idea or way of thinking about something that helps us distinguish it from other elements; can be an object, event, relationship, or process.
conceptual scheme  A set of concepts connected together to form an integrated whole that specifies and clarifies the relationships among them; individually, each concept describes a unique process; as a group, the concepts still retain common characteristics.
concurrent validity  Method for establishing criterion-related validity of a new measurement; demonstrates that the new measuring instrument and the established measuring instrument-both measuring the same or similar things-are related.
confederate  Someone who pretends to also be participating in the research project but is really helping the researcher; a deceptive practice because research participants do not know that an individual is playing the confederate role; used when the researcher needs to create a certain type of interaction context or to provide a certain type of interaction to which an unknowing research participant responds.
confessional  Writing style of the author of a qualitative research report; the author reveals everything; allows the reader to validate and verify both what was discovered and how it was discovered.
confidentiality  Protection of research participant; any information the participant provides is controlled in such a way that others do not have access to it.
confounding variable  Variable that confuses or obscures the effect of one variable on another; when present, it is often difficult to isolate the effects of independent variables.
construct  Theoretical definition of a concept; not directly observable.
construct validity  Extent to which measuring device measures the core concept that was intended to be measured and not something else; researchers use a different, but theoretically related, measure of the same or a similar phenomenon to establish construct validity.
content analysis  Quantitative research method that integrates both data collection method and analytical technique to measure the occurrence of some identifiable element in a complete text or set of messages.
content validity  Degree to which the measurement items are representative of all of the potentially items available for measuring the construct of interest.
contingency analysis  Form of chi-square in which frequency distributions are created simultaneously on two nominal variables; cases are classified on two variables in relationship to each other; also referred to as two-way chi-square or two-dimensional chi-square.
contingency table  Table in which two nominal variables are arranged; rows represent one nominal variable, and columns represent a second nominal variable; used to display data for contingency analyses.
contingent accuracy  Way to interpret meaning in qualitative research; relies on tangible artifacts, which are believed to be accurate representations of the phenomenon; most objective of the three interpretive positions.
continuous level data  Data for which values can differ in degree, amount, or frequency and for which these differences can be ordered on a continuum; also referred to as quantitative data.
control group  Group in which assigned participants receive no treatment or stimuli or receive the standard form of the treatment.
convenience sample  Sampling technique not based on random selection or probability; the researcher simply selects those who are convenient as respondents; no guarantee that all eligible units have an equal chance of being included in the sample; also referred to as an opportunity sample.
conventional pretest  Type of questionnaire pretest in which a researcher selects several individuals who are like persons in the population; the survey is completed just as it will be done in the study.
conversation analysis  Detailed analysis of a transcript of spoken discourse, the kind of normal talk people engage in every day; intended to clarify the regularities within speech and identify relationships between language and social structure.
correlation  Statistical test that examines the linear relationship between two continuous level variables; represented by the symbol r; also known as Pearson product-moment correlation coefficient.
correlation matrix  Display of variables in a table; makes it possible to see how every variable is correlated with every other variable.
criterion-related validity  Determination of whether one measurement can be linked to some other external measurement; achieved through two procedures-predictive validity and concurrent validity.
criterion variable  Variable that is influenced or changed by the predictor variable.
critical incidents  Positive or negative events remembered by participants.
Cronbach's alpha  Measure of internal reliability for a series of items across all respondents; also referred to as internal reliability or internal consistency; also known as coefficient alpha.
curvilinear  Type of relationship between two variables represented by a U-shaped curve (either concave or convex).
data  Quantitative or qualitative information about any communication phenomenon.
database  Entirety of the data from each participant or about each element compiled by variables or by cases; also referred to as a data set.
data set  Entirety of the data from each participant or about each element compiled by variables or by cases; also referred to as a database.
data triangulation  Method of triangulation in which a variety of data sources are used in one study.
debriefing  Interaction between researcher and participants immediately following the research activity; researcher explains the purpose of the study and what he or she hopes to find; any information that was withheld from participants before the research activity is shared at this time.
deception  Situation where researcher purposely misleads participants; should be used only if there is no other way to collect the data and the deception does not harm participants.
deductive  Reasoning process in which researcher begins with a theory and then gathers evidence, or data, to assess whether the theory is correct; generally used with quantitative research methods.
degrees of freedom  Number of values that vary within a statistical test; represented by the symbol df; accounts for variation due to error.
dependent variable  Variable that is influenced or changed by the independent variable.
descriptive statistics  Numbers that summarize essential and basic information about the data set as a whole.
dichotomous  Data measured at one of two nominal levels.
directional hypothesis  A precise statement indicating the nature and direction of the relationship or difference between the variables.
discourse  Set of naturally occurring messages that serve as data in qualitative methodologies.
discrete data  Data classified by the presence or absence of some characteristic or attribute; based on distinct categorical characteristics; also known as nominal data or categorical data.
discussion section  Section of the written research report in which the authors provide an interpretation of the results.
ecological validity  Form of external validity; the degree to which participants are like those the researcher is really interested in and the degree to which the research setting is natural.
empirical  Refers to observations or experiences; empirical methodologies in communication are based on or are derived from experiences with observable phenomena.
equivalent  Characteristic of nominal or categorical response set; responses must be equal to one another or of the same type.
ethnography  Detailed study and representation of people and their interaction; the holistic description of interactants in their cultural or subcultural group.
exclusion criterion  A standard or guideline that, if met, excludes participants from being selected as part of a nonprobability sample.
exhaustive  Characteristic of nominal, or categorical, response set; responses must represent the entirety of the variety of characteristics of the people or element being measured.
expected frequency  The number of times a category was expected to appear in a test of chi-square.
experiment  Research design used to determine causation; the recording of observations made by defined procedures and in defined conditions; researcher must have control over manipulation of the independent variables and random assignment of participants to conditions.
experimental research  Type of research most often conducted in the laboratory or other simulated environments controlled by researchers.
experimental variable  Variable manipulated by the researcher; presumably, this manipulation, or variation, is the cause of change in other variables; also referred to as antecedent variable, independent variable, treatment variable, and causal variable.
expert panels  Form of questionnaire pretesting in which experts in research methodology or in the survey's content read through the questionnaire together and discuss potential problems with the survey.
external validity  Degree to which the findings of a research project can be extended to participants and settings beyond those studied.
face validity  Extent to which the items reference the construct intended to be measured; exists if the measurement looks and feels as if it will capture the intended construct.
factorial design  Type of ANOVA design based on two or more categorical independent variables, each with at least two levels; allows the researcher to test for the effects of each independent variable and the interaction effect on the continuous level dependent variable.
field experiment  Form of a quasi-experimental design; like an experiment in that researcher controls the manipulation of independent variables and random assignment of participants; the research environment is realistic and natural, so researcher lacks the degree of control found in classical experiments.
field interviewing  Qualitative research method; a semidirected form of discourse or conversation with the goal of uncovering the participant's point of view.
field notes  Notes about observations made in the field, in the interaction setting or while interaction occurs.
focus group  Qualitative research method; facilitator-led group discussion used for collecting data from a group of participants about a particular topic in a limited amount of time.
frequency  Number of times a particular value or category of a variable occurs.
gatekeeper  Person who has authority to allow the researcher into an environment to collect data.
generalizability  Extent to which conclusions developed from data collected from a sample can be extended to the population; the extension of the findings to similar situations or to similar others.
heuristic  Characteristic of research; results of one study lead to more questions.
hierarchical regression  Form of regression in which the researcher determines the order or sequence in which the independent variables are presumed to influence the dependent variable.
human subjects review committee  University committee that uses a formal process for considering the soundness and reasonableness of research proposals; also known as institutional review board (IRB).
hypothesis  Tentative, educated guess or proposition about the relationship between two or more variables; often, hypotheses take the form of statements like "If x occurs, then y will follow," or "As x increases, so will y."
impressionist  Writing style of the author of a qualitative research report; intended to startle the reader; uses metaphors, phrasings, imagery, and expansive recall.
inclusion criterion  A standard or guideline participants must meet to be included in a nonprobability sample.
independent variable  Variable manipulated by the researcher; presumably, this manipulation, or variation, is the cause of change in the dependent variable; also referred to as antecedent variable, experimental variable, treatment variable, and causal variable.
inductive  Reasoning process in which data are gathered and examined, hypotheses are formulated, and eventually theories are developed in response to what the data reveal; generally used with qualitative research methods.
inferential statistics  Statistical tests that provide information about the relationships between or among variables in the study; used to draw conclusions about a population by examining the sample.
informed consent  Agreement participant gives to researcher to participate in the research project after having been given some basic information about the research process.
institutional review board (IRB)  University committee charged with the formal process of considering the soundness and reasonableness of research proposals; also known as human subjects review committee.
interaction analysis  Quantitative research method; codes the content of ongoing communication between two or more individuals; identifies the verbal or nonverbal features or functions of the stream of conversational elements.
interaction effect  Combined and simultaneous influence of two or more independent variables on the dependent variable.
intercoder reliability  See interrater reliability.
interdisciplinary triangulation  Form of triangulation in which researchers from a variety of disciplines work together on a research project.
internal reliability  Degree to which multiple items invoke the same response from a participant; expressed in value from 0 (no internal consistency) to 1.00 (complete internal consistency).
internal validity  Extent to which one can draw valid conclusions about the effects of one variable on another; depends upon how the research is designed and the data collected; addresses the relationship between the concept being measured and the process for measuring it.
interrater agreement  See interrater reliability.
interrater reliability  Degree to which two or more coders assign communication behaviors to the same categories or to which two or more raters similarly evaluate a communication act on a scale or index; also referred to as interrater agreement or intercoder reliability.
interval  Measurement based on numerical scores or values in which the distance between any two adjacent, or contiguous, data points is equal; scale without a meaningful zero.
intervening variable  Element that is presumed to explain or provide a link between other variables.
investigator triangulation  Method of triangulation in which several different researchers or evaluators are used.
justice  Issue of fairness in conducting research; addresses who should receive the benefits of research and who should bear its burdens.
key informant  Individual from whom data must be collected for a research project to be complete.
latent content  Type of content coded in content analysis; inferences or interpretations about the content that imply something about the nature of the senders or producers of the content or effects on senders.
Likert-type scale  Type of interval scale measurement widely used in communication research; participants are given a statement and then asked to respond, indicating the degree to which they agree or disagree with the statement; the response set is something like "strongly disagree, disagree, undecided, agree, strongly agree."
linear  Type of relationship between variables in which a one-unit change in one variable is associated with a constant change in the other variable in the same or opposite direction; when plotted forms a straight line.
linear regression  Form of regression in which the values of a dependent or criterion variable are attributed to one independent, or predictor, variable.
literature review  Section of the written research report that provides the framework of the research investigation; summarizes the literature the researcher sought and studied to design and develop the research study.
longitudinal  Type of research design that allows for multiple measurements of the dependent variable over time.
main effect  Simple influence of independent variable on the dependent variable; the influence of one independent variable is examined without considering the influence of other independent variables.
manifest content  Type of content coded in content analysis; a description of the characteristics of the content itself.
manipulation  One of the ways in which the researcher varies the type of stimuli or the amount or level of stimuli presented to research participants; also referred to as treatment.
manipulation check  Verification that participants did, in fact, regard the independent variable in the various ways that the researcher intended; conducted prior to statistical analyses of the hypotheses.
maturation  Threat to the internal validity of research as participants change, or mature, over the course of the observations.
maximum variation sampling  Sampling technique used in qualitative research; based on informational redundancy; a researcher continues to seek additional participants until the data received are redundant with, or the same as, previously collected data.
mean  Most common measure of central tendency; commonly referred to as the average; computed by adding up all the scores on one variable and then dividing by the number of cases, or n, for that variable.
measurement  Use of numbers to represent a communication phenomenon; more broadly, a process that includes everything the researcher does to arrive at the numerical estimates, including the measuring device or instrument, how the device or instrument is used, the skill of the person using the device or instrument, and the attribute or characteristic being measured.
median  Measure of central tendency indicating the middle of all the scores on one variable; the point or score that divides a distribution of scores in half.
methodological triangulation  Type of triangulation in which the researcher uses multiple methods-quantitative or qualitative-in one research project.
methods section  Section of the written research report that describes how the research study was executed.
mock experiment  Type of quasi-experimental research design in which a researcher creates simulated environments that differ on some variable and then observes participants' behaviors; distinguished from experimental research because a researcher loses some degree of control over the interaction.
mode  Measure of central tendency indicating the score that appears most often in a data set.
mortality  Threat to the internal validity of research when participants can no longer be observed or used for data collection because they have dropped out of the research project; also known as attrition.
multiple correlational coefficient  Represented by the symbol R; an index of the magnitude of the relationship among the variables in a multiple regression.
multiple regression  Statistic to test for the significant relationships between the dependent variable and multiple independent variables separately and as a group.
mutually exclusive  Characteristic of choices in a response set; categories should present only one option for which a person is able to identify him- or herself.
mutual simultaneous shaping  Belief that everything influences everything else in the here and now; with so much going on at once, identifying specific causality is difficult, if not impossible; used in qualitative research.
narratives  Stories people tell as a way of knowing, understanding, and explaining their lives.
negatively skewed curve  Distribution in which there are very few scores on the left side of the curve; there are very few very low scores; most of the scores are lumped together on the right side of the curve, above the mean.
network sampling  Form of nonprobability sampling in which researcher actively solicits individuals who fit a specific profile and asks them to participate in the research study.
nominal data  Discrete data that describe the presence or absence of some characteristic or attribute; data that name a characteristic without any regard to the value of the characteristic; also referred to as categorical data.
nondirectional hypothesis  Statement that a difference or relationship between variables will occur; does not specify the direction of the difference or the nature of the relationship.
non-experimental design  Type of research design in which a researcher lacks direct control over variation in the independent variables and temporal order of variables; participants are not randomly assigned to conditions; also called correlational or descriptive studies.
nonprobability sampling  Sampling technique that does not rely on any form of random selection.
nonresponse  Failure to obtain data from individuals in the sample.
normal curve  Theoretical distribution of scores, or other numerical values, in which the majority of cases are distributed around the peak in the middle with progressively fewer cases as one moves away from the middle of the distribution; has a distinct bell shape and symmetry-one side of the curve is a mirror image of the other side; the mean, median, and mode for the distribution are at the same point; also referred to as the bell curve.
null hypothesis  Implicit complementary statement to the research hypothesis that states that no relationship, except one due to chance, exists between the variables.
observed frequency  Used in chi-square; the number of times the category actually appears; is compared to the expected frequency.
observer-as-participant  Form of participant observation; researcher in both participant and observer roles, but the primary focus shifts to the role of observer, although the researcher has negotiated entrance into an interaction setting with the intention to allow the interaction to proceed as it naturally would.
one-dimensional chi-square  Statistical test to determine if differences in how the cases are distributed across the categories of one categorical, or nominal, variable are significant; also referred to as one-way chi-square.
one-tailed t-test  Statistic used to test for a specific difference on a continuous level dependent variable relative to one of two categories of a nominal level variable.
one-way ANOVA  One-way analysis of variance; statistical test to identify significant differences in the dependent variable based on categorical differences on one independent variable.
one-way chi-square  Statistical test to determine if differences in how cases are distributed across the categories of one categorical, or nominal, variable are significant; also referred to as one-dimensional chi-square.
open question  Question for which respondents use their own words to formulate a response.
operationalization  Statement that denotes how the variable is observed and measured in a specific way; most variables can be operationalized in multiple ways.
opportunity sample  Sampling technique not based on random selection or probability; the researcher selects those who are convenient to him or her as respondents; no guarantee that all eligible units have an equal chance in being included in the sample; also referred to as a convenience sample.
ordinal  Measurement based on the rank order of concepts or variables; differences among ranks need not be equal.
outcome variable  Variable that is influenced or changed by the independent, or predictor, variable.
paired comparison t-test  Statistic used to compare two paired or matched scores.
panels  Longitudinal design for a survey in which data are collected from the same participants at more than one point in time.
participant-as-observer  Form of participant observation; the researcher is both participant and observer; researcher openly acknowledges the research purpose, but also takes an active role in the interaction.
participant observation  Data collection method used with qualitative methods; allows researchers to observe communication firsthand.
Pearson product-moment correlation coefficient  Statistical test that examines the linear relationship between two continuous level variables; represented by the symbol r; also known as correlation.
percentage  Comparison between the base number and a second number; represented by the symbol %.
pilot testing  Researcher's trial of a survey or questionnaire with a small group of participants who are similar to those individuals who constitute the population before data collection actually begins; also referred to as pretesting.
planned comparisons  Hypothesized statistical comparisons used with ANOVA to compare individuals' scores on the dependent variable according to the groups or categories of the independent variable.
point biserial correlation  Form of the correlation statistic used to examine the relationship between two variables; one variable is measured at the continuous level; the other is dichotomous, or measured at one of two nominal levels.
population  All units or the universe-people or things-possessing the attributes or characteristics in which the researcher is interested.
population validity  Degree to which the sample represents the population of interest.
positively skewed curve  Distribution in which there are very few scores on the right side of the distribution or very few very high scores; most of the scores are lumped together on the left side of the curve, below the mean.
post hoc comparisons  Unplanned statistical comparisons used with ANOVA to compare individuals' scores on the dependent variable according to the groups or categories of the independent variable; conducted after the researcher finds a significant ANOVA.
posttest only  Type of research design in which participants are assigned to treatment or control groups; the simple comparison between groups allows a researcher to conclude that any significant differences found are due to the fact that the treatment group received some stimulus that participants in the control group did not.
predictive validity  Method for establishing criterion-related validity; exists when measurement predicts performance or behavior.
predictor variable  Variable that causes change in the dependent, or criterion, variable; used in non-experimental research designs because the researcher cannot directly control manipulation of the independent variable.
pretesting  Trial of a survey or questionnaire with a small group of participants who are similar to those individuals who constitute the population before data collection actually begins; also referred to as pilot testing.
pretest-posttest  Type of research design in which the dependent variable is measured before the treatment group is exposed to the stimuli; after the stimulus is given, the dependent variable is measured once again in exactly the same way with the same participants.
probability  Degree to which a particular event or relationship will occur.
probability level  Level of error the researcher is willing to accept; established for each statistical test; symbolized as p or referred to as the alpha level; also referred to as significance level.
probability sampling  Most rigorous way for identifying whom to include as part of a sample; the probability, or chance, of any element being included in the sample is known and equal for everyone or every element in the sample: also referred to as random sampling.
proprietary research  Research that is commissioned by an individual or organization for its private use.
proxy response  In responding to a questionnaire, participant makes reports about other people.
purposive sampling  Form of nonprobability sampling; depends on the judgment of the researcher who hand-picks the cases to be included in the sample; used when researcher wants to select cases that are typical of the population of interest and when sensitive topics are of research interest or when very specialized populations are sought.
qualitative methods  Research in which the researcher is the primary observer, or data collector.
quantitative data  Data that represent the values (degree, amount, or frequency) that can be ordered on a continuum; also known as continuous level data.
quantitative methods  Research that relies on numerical measurement.
quasi-experiment  Research design in which variation in the independent variable is natural, or not manipulated by the researcher; participants are not assigned randomly to treatment and control groups; also referred to as natural experiment.
questions of fact  Questions that ask for definitions for phenomena in which we are interested.
questions of policy  Questions that ask for evaluation of procedures or programs.
questions of value  Questions that ask for individuals' subjective evaluations on issues and phenomena, usually about the aesthetic or normative features of communication.
questions of variable relations  Questions that examine if, how, and the degree to which phenomena are related.
quota sampling  Form of nonprobability sampling in which a researcher uses a target, or quota, as a goal for seeking people or elements that fit the characteristics of the subgroup; when the quota for each subgroup is met, data collection is complete.
random  Characteristic of a sample in which probability for selection is equal; decreases bias.
random assignment  Procedure in which each participant or element has an equal chance of being assigned to any one of the treatment or control groups of the independent variable.
range  Simplest measure of dispersion; the value calculated by subtracting the lowest score from the highest score.
ratio data  Measurement for which intervals between data points are equal; a true zero exists; if the score is zero, there is a complete absence of the variable.
raw data  Data in the form in which they are collected from each participant or about each element; compiled with data from all participants into a data set.
realist  Writing style for a qualitative research report; author narrates the story in a dispassionate, third-person voice.
recall cue  Statement preceding a survey or questionnaire designed to direct participants to recall episodes or past interactions in which they participated.
reference list  Section of the written research report that provides an alphabetical listing by authors' last names of all materials cited or referenced in the research report.
regression  Set of statistical techniques that predict some variables by knowing others; the most common use of regression is to assess the influence of several continuous level predictor, or independent, variables on a single continuous criterion, or dependent, variable.
regression line  Line drawn through the data points on a scattergram that best summarizes the relationship between the independent and dependent variables (or the predictor and criterion variables).
reliability  Achieved when researchers are consistent in their use of data collection procedures and when participants react similarly to them; other researchers using the same measure in another project with comparable participants would produce similar results; measurement is stable, trustworthy, or dependable; a reliable measure is one that is consistent or gives very similar results each time it is used.
reliability coefficient  Number between 0 and 1 to express the degree of consistency in measurement; 1.00 represents complete consistency or reliability; a score of 0 indicates that consistency or reliability was no better than would be due to chance.
repeated measures  Form of ANOVA design in which each participant is measured more than once, usually at different levels or for different conditions; also referred to as within-subject design.
replication  Design of current study to follow the procedures of other studies that have investigated the same topic, often using the same methods or procedures.
research  The discovery of answers to questions through the application of scientific and systematic procedures.
researcher construction  Way in which researcher interprets meaning in qualitative research; interpretation from researcher's personal, subjective position or perspective on the experience; evidence is fully the construction of the researcher.
research protocol  Written detailed procedures for conducting the research study and collecting data.
research question  Question that asks what the tentative relationship among variables might be or asks about the state or nature of some communication phenomenon.
respect for persons  Two separate principles: treating individuals as capable of making decisions and protecting those who are not capable of making their own decisions.
respondent validation  Researcher asks participants to review research notes or interpretations.
response rate  Number of people who respond after they have been contacted as part of the sample and asked to participate; divide the number of people who responded by the number of respondents identified as part of the sample; also referred to as return rate.
results section  Section of the written research report that provides the results of the study without interpretation.
rho  The statistic for Spearman correlation coefficient; examines the relationship between two variables, both with data captured on ordinal scales.
sample  A subset, or portion, of a population; data are collected from a sample to make generalizations back to a population.
sample size  Number of people or elements from whom or on which data are collected.
sampling error  Degree to which a sample differs from population characteristics on some measurement; as sample size increases and becomes a larger proportion of the population, sampling error is reduced; also referred to as the margin of error.
sampling frame  Set of people or elements that are available to be selected as part of the sample; the list of the available population from which participants are selected.
self-administered survey  Data collection method in which a participant reads and selects a response without the researcher's aid; also called self-report.
self-report  Type of survey, questionnaire, or poll in which respondents read the question and select a response by themselves without researcher interference.
semantic differential scale  Form of interval measurement; using a stimulus statement, participants are asked to locate the meaning they ascribe to the stimulus on a response scale anchored by two opposites, usually bipolar adjectives.
significance level  Level of error the researcher is willing to accept; established for each statistical test; symbolized as p or referred to as the alpha level; also referred to as probability level.
simple random sample  Sampling technique in which every person or unit-selected one at a time and independently-has an equal chance of being selected to participate in the study.
simulation  Type of quasi-experimental research design in which a researcher creates a research environment as close as possible to the natural environment of interest.
skewed distribution  Shape of a distribution of scores that is not normal; the curve is asymmetrical; the mean, median, and mode will not be at the same point.
snowball sampling  Nonprobability sampling technique in which participants help the researcher identify other similar participants; used when the research topic is controversial or a specific population of participants is difficult to find.
social desirability response  Response for which there is the potential for participants to respond with answers they believe the interviewer will perceive as favorable.
social science research  Research conducted through the use of scientific and systematic methods; based on the assumption that research can uncover patterns in the lives of people.
Spearman correlation coefficient  A statistic, rho, used to examine the relationship between two variables, both with data captured on ordinal scales.
split-half reliability  Expression of internal consistency between two separate but equal versions of the same test or questionnaire.
sponsor  Someone to vouch for person in the role of researcher; also validates and legitimizes researcher's presence.
spurious correlation  Relationship between two variables in which a third variable-sometimes identified, at other times unknown-is influencing the variables tested. Also called spurious relationship.
standard deviation  Representation of the variability or spread of the data set; the amount the scores in a distribution deviate from the mean.
statistical inference  Accepting the conclusions derived from the sample and assuming that those conclusions are also applicable to the population; based on the probability level computed for each statistical test.
stepwise regression  Form of regression in which independent variables are entered in sequence or order as determined by the statistical program based on the degree of influence the independent variables have on the dependent variables.
stratified random sampling  Form of random or probability sampling in which the population is divided according to subgroups of interest, or homogeneous groups; then elements are randomly selected from each homogeneous subgroup of the population with respect to its proportion to the whole.
subjective valuing  Way in which researcher interprets meaning in qualitative research; interpretation relies on a mix of both objective and subjective elements; the researcher mixes his or her interpretations with interpretations received directly from participants.
subjectivity  Approach to research in which researcher uses interpretive research processes to make the subject of the interpretation meaningful.
survey  System for collecting information to describe, compare, or explain knowledge, attitudes, and behavior; also known as questionnaire or poll.
survey objective  Statement of outcomes the researcher hopes to obtain by using a survey.
systematic sampling  Form of random or probability sampling in which every nth element is chosen after starting at a random point.
test-retest reliability  The expression of the relationship, or correlation between scores at two administrations of the same test, or measurement, to the same participants; 1.00 would indicate that the same participants evaluated the stimulus the second time exactly as they did the first time; a score of 0 would indicate that the two measurements were no better related than they would be by chance.
theory  Related set of ideas that explains how or why something happens; presents a systematic view of the phenomenon; specifies the relationships among the concepts with the objective of describing, explaining, and predicting the phenomenon.
theory triangulation  Method of triangulation; occurs when a research project uses multiple perspectives or multiple theories to interpret a single set of data; researchers attempt to demonstrate the validity of one theoretical explanation over another.
treatment  One of the ways in which the researcher varies the type of stimuli, or the amount or level of stimuli, presented to research participants; also referred to as manipulation.
treatment group  The group of participants who receive a stimulus, or one manipulation of the independent variable.
treatment variable  Variable manipulated by the researcher; presumably, this manipulation, or variation, is the cause of change in the dependent variable; also referred to as antecedent variable, experimental variable, independent variable, and causal variable.
triangulation  Use of several kinds of methods or data to further validate research outcomes and results.
t-test  Statistic used to find differences between two groupings of the independent variable on a continuous level dependent variable.
two-dimensional chi-square  Form of chi-square in which frequency distributions are created on two nominal variables; cases are classified on two variables in relationship to each other; also referred to as contingency analysis or two-way chi-square.
two-tailed t-test  Statistic used to test for differences on a continuous level dependent variable relative to one of two categories of a nominal level variable; no specific difference identified in hypothesis or research question.
two-way ANOVA  Statistical test to examine the influence of two categorical independent variables on one continuous level dependent variable; can determine the main effect of contributions of each independent variable and the interaction effect.
two-way chi-square  Form of chi-square in which frequency distributions are created on two nominal variables; cases are classified on two vari-ables in relationship to each other; also referred to as contingency analysis or two-dimensional chi-square.
Type I error  Error in hypothesis testing that occurs when the null hypothesis is rejected even when it is true; set or controlled by the researcher when choosing the significance level for the statistical test.
Type II error  Error in hypothesis testing that occurs when the alternative hypothesis is rejected even when it is true.
unitizing reliability  Degree to which two or more observers agree that communication action has taken place and can identify the interaction in the stream of interaction; used in content analysis.
unit of analysis  Discrete element coded and counted in content analysis; an observable and measurable unit that provides a standard way of dissecting the text or content into elements to be analyzed.
validity  Achieved when the measurement does what it is intended to do; related to truthfulness or accuracy in measurement.
variable  Element that is specifically identified in the research hypotheses or questions; must be able to be expressed as more than one value or in various categories.
variance  Dispersion of the distribution of scores.
within-groups variance  Variation among scores within any category or level of an independent variable.
within-subject design  Form of ANOVA design is which each participant is measured more than once, usually at different levels or for different conditions; also referred to as repeated measures design.