1/81
Comprehensive vocabulary flashcards covering survey methods, sampling designs, content analysis, descriptive and inferential statistics, threats to validity, experimental designs, and research report writing from the Exam 2 Study Guide.
Open-Ended Question
A survey question format in which participants generate their own responses, providing greater detail and allowing for unforeseeable answers, but requiring time-intensive coding.
Close-Ended Question
A survey question format where respondents select from predetermined options, allowing uniform responses, easy comparability, and quick analysis, but potentially forcing choices that respondents do not agree with.
Social Desirability Bias
The tendency of survey respondents to over-report socially approved behaviors and underreport undesirable behaviors, which can be mitigated by ensuring anonymity and phrasing questions to save face.
Cross-Sectional Survey
A survey administered at a single point in time with the goal of using the captured data as the entire basis for the study.
Trend Longitudinal Survey
A longitudinal survey design that draws a new sample meeting the same inclusion criteria from the general population each time data is gathered, meaning individuals eligible at one wave may or may not be eligible at subsequent waves.
Cohort Longitudinal Survey
A longitudinal survey administered multiple times to an unchanging pool of eligible participants, meaning that if an individual is eligible once, they remain permanently eligible across all administrations.
Panel Longitudinal Survey
A longitudinal study design in which the exact same sample of individuals surveyed at the initial measurement point is re-surveyed at multiple subsequent points in time.
Absolute Frequency
An ideal, ratio-level measure of frequency that specifies the exact, numeric count of occurrences.
Relative Frequency
A less specific, ordinal-style measure of frequency indicating how often an event occurs using qualitative categories such as "most always" or "almost never."
Skip Pattern
A survey questionnaire design element that routes respondents to specific follow-up questions or permits them to bypass irrelevant sections based on their prior answers.
Order Effects
The bias or alteration in responses that occurs when answering an earlier question influences how a respondent interprets or answers subsequent questions.
Push Poll
A controversial political tactic disguised as an objective telephone poll, designed to influence the respondent's views and public perception using biased or false statements.
Probability Sampling
A selection process where every member of a population has a known, non-zero chance of selection, permitting statistical generalization to the broader population.
Non-Probability Sampling
Any selection method where members of a target population do not have an equal or known probability of selection, meaning findings generally cannot be statistically generalized.
Census
A comprehensive study that measures every single member of an entire defined population.
Margin of Error
A statistical estimate expressing how closely the observed sample results match the true values in the underlying population.
Confidence Level
The degree of statistical certainty that a population parameter falls within the calculated margin of error.
Universe of Messages
The complete collection of all communication messages that meet a researcher's defined criteria when the primary research interest is in textual or media content.
Power Analysis
A statistical calculation (cited by Wiedmaier) that uses desired effect size, level of measurement, and statistical test type to determine the minimum sample size required to detect meaningful comparisons.
Systematic Random Sampling
A probability sampling procedure where a researcher numbers all list members, selects a random starting element, and then selects every kth element by dividing population size by sample size.
Stratified Random Sampling
A probability sampling method where a population is divided into non-overlapping groups (strata) based on important characteristics to ensure those subgroups are proportionally represented.
Multi-Stage Cluster Sampling
A sampling approach used when a complete population list is unavailable, involving the random selection of naturally occurring groupings (such as zip codes) followed by random selection of individuals within those clusters.
Purposive Sampling
A non-probability technique (also called judgmental sampling) where specific individuals or specialized subsets of a population (such as substance users) are intentionally chosen based on known characteristics.
Quota Sampling
A non-probability version of stratified sampling in which individuals are gathered non-randomly until pre-set target percentages or counts for specified subgroups (e.g., 50% female, 50% male) are filled.
Snowball Sampling
A non-probability referral method where initial participants recruit or refer acquaintances to join the study, particularly valuable for studying hidden networks or hard-to-reach populations.
Content Analysis
A research technique for systematically and quantitatively examining the characteristics of recorded communication.
Manifest Content
Surface-level, countable, and directly observable communication elements whose presence or absence is easily determined without deep interpretation (e.g., counts of male and female actors).
Latent Content
Underlying, subjective meanings, themes, or symbolic messages embedded within communication content (e.g., portrayals of friendship).
Sentiment Analysis
A form of content analysis (also called tonality analysis) evaluating the valence—positive, negative, or neutral—of a communicator's attitude toward a subject.
Prominence Analysis
A content analysis assessment evaluating media items based on the publication's reputation, article length, and location/prominence within the print or broadcast layout.
Social Media Share of Voice
An exposure metric calculated by dividing an organization's total social media mentions and references by the combined total mentions across all competitors in the market.
Customer Satisfaction Score (CSat)
A formal advocacy metric captured via website pop-up prompts prompting users to rate their satisfaction immediately following an interaction.
Net Promoter Score (NPS)
A standardized customer loyalty metric that queries respondents on a scale from 0 ("not at all likely") to 10 ("extremely likely") regarding how probable they are to recommend a company or service to friends.
Intercoder Reliability
The degree of coding consistency and agreement exhibited when multiple independent coders evaluate the exact same content using a coding guide.
Intracoder Reliability
The degree of consistency and agreement demonstrated when the same individual coder categorizes the same content at separate intervals over time.
Central Tendency
A category of descriptive statistics describing the central location or middle point of a distribution where data cluster together.
Dispersion
Descriptive statistics reflecting how scores vary, differ, and spread out relative to the center of a distribution.
Skewness
An asymmetry in frequency distribution characterized by an excess of scores clustering at one end and a trailing tail stretching toward the opposite extreme.
Mean
The arithmetic average calculated by summing all observations and dividing by the total count; requires interval or ratio measurement and is vulnerable to distortion by outliers.
Median
The exact middle value dividing the upper 50% of scores from the lower 50%; requires at least ordinal data and is determined by averaging the two central values when the sample count is even.
Mode
The single score or category that occurs most frequently in a dataset; the only measure of central tendency usable on nominal data.
Bimodality
A distribution possessing two distinct, high-frequency peaks at opposing ends, making the mean misleading and requiring the reporting of modes.
Range
A basic dispersion metric representing the arithmetic difference between the maximum and minimum observations in a data set.
Total Deviation
The sum of squared deviances from the mean; squaring each raw deviation prevents negative and positive values from canceling each other out to zero.
Variance
A dispersion statistic calculating the mean of the squared deviances from the distribution average; accounts for sample size differences but is expressed in squared units.
Standard Deviation
The square root of variance, expressing the average distance of scores from the mean in the original metric of measurement.
Spurious Correlation
An observed, statistically significant correlation between two variables that lack a true causal connection, typically generated by chance or a shared, unmeasured third variable.
Inferential Statistics
Statistical procedures analyzing two or more variables simultaneously to test hypotheses and establish whether observed patterns can be reliably generalized to a broader population.
Null Hypothesis
The formal baseline assumption in hypothesis testing asserting that no relationship or difference exists between the examined variables in the population.
Alternative Hypothesis
The researcher's operational hypothesis asserting that a real effect, difference, or relationship exists; classified as either directional (one-tailed) or non-directional (two-tailed).
p-Value
The probability that the observed sample difference or correlation would occur by random chance if the null hypothesis were true; in social sciences, p<.05 is the accepted threshold for statistical significance.
Type I Error
A false positive decision occurring when a researcher incorrectly rejects a null hypothesis that is actually true; the risk is capped by the chosen alpha level (p<.05).
Type II Error
A false negative decision occurring when a researcher fails to reject a null hypothesis that is genuinely false, failing to identify an effect that is truly present.
Pearson's r
A parametric correlation coefficient used to measure the linear strength and direction of association between two continuous, quantitative variables.
Analysis of Variance (ANOVA)
A parametric statistical test that evaluates differences across two or more group means by comparing the variance between groups to the variance within groups.
Post-Hoc Test
Follow-up pairwise statistical tests conducted after an omnibus ANOVA reaches significance across three or more groups, pinpointing exactly which specific pairs of group means differ.
Chi-Square Test
A nonparametric cross-tabulation statistic used to evaluate whether observed distributions across two categorical, nominal variables differ significantly from expected frequencies.
Statistical Power
The probability that a statistical test will correctly detect a genuine difference when one exists; directly influenced by sample size according to Hocking et al.
Internal Validity
The degree of confidence that changes in the dependent variable are exclusively attributable to the manipulated independent variable rather than uncontrolled extraneous variables.
External Validity
The extent to which research conclusions can be legitimately generalized across other settings, real-world contexts, and distinct target populations.
History
A threat to internal validity arising when an unplanned external event occurs during the study and alters participant performance on the dependent variable.
Maturation
A threat to internal validity resulting from natural internal biological, cognitive, or physical changes occurring in subjects over elapsed time.
Testing / Sensitization
A threat to internal validity caused when completion of an initial pre-test increases participant familiarity or sensitivity, artificially altering their post-test scores.
Regression to the Mean
A statistical tendency where participants selected due to extreme scores on a first testing occasion naturally score closer to the overall group average upon repeated testing.
Experimental Mortality
A threat to internal validity stemming from participant attrition, where subjects discontinue or drop out of an experimental study before its conclusion.
Contamination
A threat to validity occurring when participants who have completed a study disclose details of experimental procedures or objectives to future subjects.
Random Assignment
The allocation of study participants into treatment or control conditions purely by chance, ensuring initial group equivalence and neutralizing selection bias.
Post-Test Only Control Group Design
A true experimental setup where randomly assigned experimental and control groups are compared exclusively after treatment administration, limiting pre-test sensitization but leaving baseline equality unverified.
Pre-Post Control Group Design
An experimental framework in which both treatment and control groups receive a baseline pre-test and a final post-test, confirming initial equivalence but introducing vulnerability to sensitization.
Solomon Four-Group Design
A sophisticated experimental design using four groups (two with pre-tests and two without) to simultaneously determine treatment impact while isolating any confounding effect of pre-test sensitization.
Repeated Measures Design
A within-group experimental configuration where each individual participant completes every condition, acting as their own baseline control.
Counterbalanced Design
A procedural control applied in repeated-measures studies that varies the order in which conditions are presented across participants to neutralize order effects.
Factorial Design
An experimental framework incorporating two or more independent variables simultaneously, allowing the analysis of main effects and interaction effects between variables.
Quasi-Experiment
An empirical study design that mirrors an experiment by manipulating an independent variable but lacks random assignment, relying instead on pre-existing or intact groups.
Interrupted Time-Series Design
A quasi-experimental structure that collects repeated observations across several intervals before and after the introduction of an intervention.
Reactivity
The alteration of behavior displayed by experimental participants driven solely by their awareness that they are being observed or evaluated.
Extraneous Variable
Any unmeasured variable unrelated to the study hypothesis that possesses the potential to influence the dependent variable, originating from participants, settings, or researchers.
Confounding Variable
An uncontrolled factor that covaries systematically with the independent variable, preventing researchers from isolating which factor caused changes in the dependent variable.
Ecological Validity
A subtype of external validity addressing whether findings obtained within tightly controlled laboratory conditions can be generalized to naturalistic, real-world environments.
Population Validity
A component of external validity evaluating whether the experimental sample adequately reflects the larger population to support valid generalization.
Abstract
A brief summary of an academic research article typically between 150–250 words, accompanied by relevant key words.
Executive Summary
A high-level overview written for business decision-makers, typically encompassing approximately 10% of a final report's length, summarizing problem background, research objectives, methodology, key findings, and practical recommendations.