Research Methods Midterm Ultimate Study Guide

Criteria for Causality and Experimental Design

• Three Essential Requirements for Claiming Causality:

  • Covariance / Correlation: There must be a statistically demonstrated relationship or association between the claimed cause (Independent Variable) and the claimed effect (Dependent Variable).

  • Temporal Precedence: The cause must precede the effect in time. The Independent Variable ($IV$) must be manipulated or occur prior to the observed change in the Dependent Variable ($DV$).

  • Internal Validity / Elimination of Confounds: All alternative explanations and potential confounding variables must be systematically ruled out or controlled.

• Bike Scenario Methodological Evaluation:

  • Independent Variable ($IV$): The variable being manipulated (e.g., the specific intervention or bike adjustment).

  • Dependent Variable ($DV$): The outcome measure being evaluated (e.g., speed, balance, or completion time).

  • Scientific Improvements Needed: To establish a true causal claim, the setup requires a rigorous experimental control group, elimination of environmental confounds, and systematic random assignment.

• Random Assignment:

  • Definition: A core experimental procedure wherein every single participant in a study has an equal mathematical chance of being placed into any of the experimental conditions or control groups.

  • Implementation Methods: Executed using objective randomization tools such as coin tosses, random number generators/tables, or lotteries.

  • Function: With a sufficiently large sample size, random assignment ensures that all participant groups are equivalent across all individual difference dimensions (e.g., baseline trait levels, intelligence, personality, demographics) prior to the administration of the Independent Variable.

Priming and Cognitive Flexibility Studies

• Bargh Conceptual Priming Experiment Case Study:

  • Hypothesis: Activating specific social concepts implicitly alters subsequent social interactions and behavior.

  • Independent Variable ($IV$): Prime type across three conditions: Neutral Prime, Rude Prime, and Polite Prime.

  • Dependent Variable ($DV$): Behavioral outcome measured as whether or not the participant interrupted the experimenter within a $10$\,minute observation window.

  • Experimental Control Procedures:

  • Tasks and experimental locations were kept identical across all conditions.

  • The task consisted of a scrambled sentence completion test, differing solely in the specific target prime words embedded within the materials.

  • Both the experimenter and the research confederates were kept double-blind to the participant's assigned experimental condition.

  • Experimental Stimuli / Materials:

  • Rude Prime Condition: Sentence scrambles containing rude concept words (e.g., "they her bother see usually").

  • Polite Prime Condition: Sentence scrambles containing polite concept words (e.g., "they her respect see usually").

  • Neutral Prime Condition: Sentence scrambles containing neutral concept words (e.g., "they her send see usually").

  • Empirical Results (Percentage of Participants Who Interrupted):

  • Rude Prime Condition: Approximately $60\%$ to $70\%$ interrupted.

  • Control / Neutral Condition: Approximately $38\%$ interrupted.

  • Polite Prime Condition: Approximately $18\%$ to $20\%$ interrupted.

  • Causal Conclusion: The explicit manipulation of the independent variable (activating concepts of rudeness versus politeness versus neutrality) directly caused the observed variance in the dependent variable (the probability of interrupting the experimenter).

• Real-Time Strategy Video Games and Brain Flexibility Study:

  • Citation & Source Details: WebMD News from HealthDay by Robert Preidt (Published Friday, August 23, 2013). Research led by Dr. Brian Glass of the School of Biological and Chemical Sciences at Queen Mary University of London.

  • Main Finding: Playing specific real-time strategy (RTS) video games promotes flexible thinking skills, honing the cognitive ability to think on the fly and adaptively learn from past mistakes.

  • Clinical Implications: Results offer potential frameworks for designing novel cognitive rehabilitation treatments for individuals suffering from traumatic brain injuries or neurodevelopmental conditions such as Attention-Deficit/Hyperactivity Disorder (ADHD).

  • Distinction from Prior Research: While action video games speed up immediate decision-making response times, real-time strategy games specifically build tactical flexibility and rapid resource management under shifting conditions.

  • Sample Demographics & Methodology:

  • Participants: $72$ female adult participants who typically played video games for less than $2$\,hours per week at baseline. The study authors explicitly noted that male gamers could not be included because they were unable to locate any male participants who played less than $2$\,hours per week.

  • Condition 1 (Experimental Groups): Two-thirds ($2/3$) of the sample played either basic or complex versions of the real-time strategy game "StarCraft" (a fast-paced tactical game requiring players to construct, organize, and command armies to battle enemies).

  • Condition 2 (Control Group): One-third ($1/3$) of the sample played "The Sims" (a life simulation game that does not rely on memory retention, speed, or tactical strategic planning).

Methodological Trade-Offs and Combining Approaches

• Barriers to Executing Pure Experiments:

  • Ethical & Feasibility Constraints:

  • It is unethical to manipulate harmful or traumatic life events as independent variables (e.g., death of a spouse, placing people in incompatible romantic partnerships, or inducing disease and physical injury).

  • It is physically or practically impossible to manipulate static biological characteristics (e.g., biological sex, ethnicity, innate genetic traits).

  • Frequency Goals: When research aims to answer descriptive population questions (frequency claims), laboratory manipulations introduce unnecessary artificiality. In frequency claims, isolating mechanisms is less critical than capturing authentic real-world manifestations.

  • Artificiality Concerns: Laboratory settings are contrived and constrained, often failing to mirror real-world context. However, moving research into real-world settings reduces experimental control and prevents the total elimination of confounds.

  • Effect Size Discrepancies: Laboratory effect sizes may differ from real-world effect sizes because real-world environments contain significantly greater measurement noise and extraneous error variance.

• Rebuttal for Experimental Validity:

  • Empirical meta-analyses demonstrate that laboratory and field studies typically yield consistent, highly similar findings.

  • Distinction between Mundane Realism (the extent to which the physical setting mimics real-world environments) and Psychological Realism (the extent to which the psychological processes triggered in the lab are authentic to real-life cognitive processes).

• Multi-Method Research Programs:

  • Researchers systematically combine methodologies to optimize internal and external validity:

  • Combining non-experimental (correlational/observational) and experimental designs.

  • Conducting paired field studies and laboratory experiments.

  • Integrating behavioral tracking with neuroscience methods (e.g., EEG, fMRI, EMG, GSR).

Navigating Psychological Literature and Peer Review

• The Academic Peer Review Cycle:

  • Step 1: Researchers design a study, collect and analyze data, and draft a formal manuscript.

  • Step 2: The manuscript is submitted to a specialized academic journal.

  • Step 3: The journal editor forms an initial impression and either desk-rejects or sends the paper to $2$ to $4$ expert peer reviewers in the field.

  • Step 4: Over a review period of $1$ to $6$\,months, reviewers evaluate the manuscript double-blind (the authors are blind to reviewer identities).

  • Step 5: The editor writes a decision letter to the authors outlining one of three outcomes:

  • Accept: The manuscript is accepted as is and classified as "in press" until publication.

  • Revise and Resubmit (R&R): Authors must address reviewer critique, make revisions, and submit a revised manuscript with a point-by-point cover letter.

  • Reject: The manuscript is declined for publication.

• Strategies for Reading Journal Articles efficiently:

  • Start with the Abstract: Use the abstract as a foundational mental blueprint. Re-read the abstract whenever dense material creates confusion.

  • Ignore Low-Level Statistical Details Initially: Avoid getting bogged down in complex numerical calculations during early passes; focus on understanding the primary statistical conclusion.

  • Locate Key Conceptual Information:

  • Hypotheses and theoretical background are located in the final paragraph(s) of the Introduction.

  • Summaries of main findings and interpretations are located in the first paragraph(s) of the Discussion.

  • Structured Note-Taking Questions:

  • What was the authors' main hypothesis?

  • Why did they predict this outcome theoretically?

  • How did they operationalize and test this hypothesis?

  • What were the key empirical findings?

  • What do these findings mean conceptually?

  • What are your critical reactions (e.g., surprising trends, inconsistencies with prior knowledge)?

• "Forest Then Trees" Perspective:

  • Big Picture Focus: Identify the central point the authors are conveying and the novel knowledge contributed to the literature.

  • Critical Validity Evaluation: Critically evaluate the research across Construct, External, Statistical Conclusion, and Internal Validity based on the claim type.

  • Core Psychological Question: Always evaluate articles by asking: "What does this tell me about how people think, feel, or act in general?"

  • Synthesis Value: Summarize the core contribution in $1$ or $2$ concise sentences. Single-sentence big-picture syntheses provide more impactful academic citation support than paragraph-long listings of granular low-level details.

Data Visualization: Reading Tables and Graphs

• Identifying Relevant Variables in Data Results:

  • Search text and displays for Independent Variables ($IV$) and Dependent Variables ($DV$).

  • Search for key operational phrases: manipulated/assigned, significant, as hypothesized/predicted, consistent with predictions, greater than / less than.

• Core Statistical Metrics:

  • Mean ($M$): Arithmetic average score.

  • Standard Deviation ($SD$): Measure of score dispersion around the mean.

  • $p$-value ($p$): The calculated probability that the observed effect occurred by chance alone under the null hypothesis.

  • If $p < .05$: The effect is statistically significant; reject the null hypothesis.

  • If $p > .05$: The effect is non-significant; fail to reject the null hypothesis.

• Reading Data Tables:

  • Table Title: Defines what the numbers inside the cells represent (typically the Dependent Variable).

  • Table Headings: Categorize columns and rows, representing levels of the Independent Variable(s) or measured grouping factors.

  • Table Case Studies and Empirical Breakdown:

  • Self-Consciousness and Task Performance Study:

    • Title indicates mean performance scores (higher scores = better performance).

    • Manipulated $IV$ = Task Focus (Hands focus vs. Control); Measured Variable = Self-consciousness (High vs. Low).

    • High Self-consciousness + Hands Focus: Adjusted Mean = $49.4$.

    • High Self-consciousness + Control Focus: Adjusted Mean = $59.6$.

    • Low Self-consciousness + Hands Focus: Adjusted Mean = $52.2$.

    • Low Self-consciousness + Control Focus: Adjusted Mean = $71.7$.

  • Implementation Intentions and Daily Snacking Study:

    • Descriptive Statistics: Participants consumed an average of $1.7$\,healthy snacks/day ($SD = 1.2$; $73\%$ composed of fruits and vegetables) and an average of $209$\,kcal/day from unhealthy snacks ($SD = 147$). Baseline motivation to eat healthily was high ($M = 5.2$, $SD = 1.0$) and did not differ across conditions ($F(2, 105) = 0.43, p = .65$).

    • MANOVA Findings: Significant multivariate main effect of condition on overall snack intake (Wilks's Lambda $F(4, 208) = 3.08, p < .05$).

    • Univariate Analyses: Condition effect was statistically significant for healthy snacking ($F(2, 105) = 5.84, p < .01$), but non-significant for unhealthy snacking ($F(2, 105) = 0.20, p = .82$).

    • Condition Means Breakdown:

    • Implementation Intention + Situational Cue ($n = 45$): Healthy snacks $M = 1.6$ ($SD = 1.1$); Unhealthy snacks $M = 217$\,kcal ($SD = 141$).

    • Implementation Intention + Motivational Cue ($n = 39$): Healthy snacks $M = 2.1$ ($SD = 1.4$); Unhealthy snacks $M = 207$\,kcal ($SD = 156$).

    • Control Condition ($n = 24$): Healthy snacks $M = 1.1$ ($SD = 0.7$); Unhealthy snacks $M = 194$\,kcal ($SD = 140$).

  • Mood Regulation, Mood Freeze, and Snacking Study:

    • Evaluated hypothesis that people snack during negative mood states because they believe eating improves mood.

    • Manipulated $IV_1$ = Mood (Happy vs. Distress); Manipulated $IV_2$ = Believed Mood Changeability (Changeable Mood vs. Mood Freeze).

    • Outcome ($DV$) = Sum of standardized amounts of cookies, pretzels, and cheese crackers eaten (positive scores indicate increased eating).

    • Happy + Mood Freeze: $M = 0.49$ ($SD = 1.78$).

    • Happy + Changeable Mood: $M = -0.35$ ($SD = 2.30$).

    • Distress + Mood Freeze: $M = -0.89$ ($SD = 1.62$).

    • Distress + Changeable Mood: $M = 0.79$ ($SD = 3.00$).

    • Interpretation: Distressed participants only increased snack consumption when they believed their mood could change, supporting the mood-regulation theory.

  • Social Rejection and Aggression Noise-Blasting Study:

    • Manipulated $IV$ = Social Rejection (Accepted vs. Rejected).

    • Aggression Composite: Accepted $M = -0.94$ ($SD = 1.03$); Rejected $M = 0.94$ ($SD = 1.76$); $F(1, 28) = 12.76, p < .01$ (Significant).

    • Noise Intensity Choice: Accepted $M = 3.53$ ($SD = 1.88$); Rejected $M = 6.60$ ($SD = 2.20$); $F(1, 28) = 16.83, p < .001$ (Significant).

    • Noise Duration: Accepted $M = 771.20$ ($SD = 745.84$); Rejected $M = 1,658.80$ ($SD = 1,610.36$); $F(1, 28) = 3.75, p > .05$ (Non-significant).

    • Positive Mood: Accepted $M = 27.27$ ($SD = 9.11$); Rejected $M = 25.73$ ($SD = 7.19$); $F(1, 28) = 0.26, p > .05$ (Non-significant).

    • Negative Mood: Accepted $M = 13.27$ ($SD = 3.01$); Rejected $M = 14.33$ ($SD = 4.72$); $F(1, 28) = 0.55, p > .05$ (Non-significant).

• Reading Graphical Data:

  • Axes Convention: $Y$-axis displays the Dependent Variable ($DV$) or Frequency; $X$-axis displays the Independent Variable ($IV$) or Predictor Variable.

  • Graphical Case Studies:

  • Chewing Gum Craving Study Interaction Line Graph:

    • Evaluated craving scores across Time 1 and Time 2 for Chewing Gum vs. No-Gum conditions.

    • Significant Condition $\times$ Time interaction observed: $F(1, 18) = 13.36, p < .01$.

    • Simple Effects Tests: Non-significant difference at Time 1 ($F(1, 18) = 0.21, p > .05$); statistically significant difference at Time 2 ($F(1, 18) = 38.04, p < .01$) showing substantial craving reduction in the Gum condition.

  • Procrastination and Symptom Reporting Line Graph:

    • Categorized participants as Procrastinators vs. Nonprocrastinators based on a median split of Lay's General Procrastination Scale ($M = 42.7$, Median = $45$, Range = $18$ to $63$).

    • Early Semester Health Symptoms: Nonprocrastinators reported $M = 2.8$\,symptoms/week; Procrastinators reported $M = 1.4$\,symptoms/week.

    • Late Semester Health Symptoms: Nonprocrastinators reported $M = 5.2$\,symptoms/week; Procrastinators reported $M = 8.2$\,symptoms/week.

Scientific Claims and the Four Validities

• Types of Relationships Observed in Scatterplots & Visuals:

  • Positive Association: Higher values on Variable $X$ correspond to higher values on Variable $Y$.

  • Example: National GDP Per Capita vs. Life Satisfaction (Our World in Data 2018; higher national income correlates with higher life satisfaction in nations like Finland, Netherlands, Switzerland, and Norway).

  • Negative Association: Higher values on Variable $X$ correspond to lower values on Variable $Y$.

  • Example 1: Average avocado price vs. total weekly avocado sales across $54$ US cities (sales drop from $150,000,000$ to $75,000,000$ as average price increases from $0.75$ to $1.50$ dollars; sales spike sharply around the Super Bowl as prices drop).

  • Example 2: Beck Depression Inventory-II (BDI-II) scores vs. Optimistic/Pessimistic Bias ($r = -.35, p < .0001, N = 153$; Low BDI $M = .01$, Middle BDI $M = -.02$, High BDI $M = -.06$).

  • No Association: Changes in Variable $X$ show no systematic linear relationship with Variable $Y$ (e.g., frequency of prescribed ADHD drug use vs. rate of illicit drug use).

  • Curvilinear Association: The relationship pattern flips direction across levels of Variable $X$.

  • Example: Anxiety level vs. Performance (Yerkes-Dodson model): Low anxiety yields low performance, moderate anxiety yields optimal peak performance, and high anxiety impairs performance.

  • Categorical Comparisons (Bar Graphs):

  • College vs. Alternative Investments Internal Rate of Return (IRR): Associate's Degree ($20\%$ IRR) and Bachelor's Degree ($15\%$ IRR) yield higher inflation-adjusted returns than Stocks ($\sim 7\%$), Corporate Bonds ($\sim 3\%$), Gold, Treasury Bills, and Housing ($< 1\%$).

  • Netspeak Keyword Impact on Online Messaging Reply Rates: Base average reply rate is $32\%$. Keywords such as "hit", "ur", "r", "u", "ya", "cant" drag response rates down below $10\%$, whereas "wat", "realy", and "luv" maintain relatively higher reply rates.

• The Four Four Core Validities:

  • Construct Validity: How well an operationalized variable represents the abstract theoretical definition.

  • Internal Validity: The degree of certainty that changes in the Independent Variable ($IV$) directly caused changes in the Dependent Variable ($DV$), ruling out all confounds.

  • External Validity: The extent to which findings generalize to other populations, settings, operationalizations, cultures, and time periods.

  • Statistical Conclusion Validity: The degree to which statistical conclusions are precise and reasonable ($p < .05$, margin of error, effect sizes).

• Mapping Validities to Scientific Claim Types:

  • Frequency Claims: State a single rate or percentage regarding one measured variable.

  • Validities Assessed: Construct Validity, Statistical Conclusion Validity (margin of error, sample size), External Validity. Internal Validity is not applicable.

  • Case Study 1: "$36\%$ of millennials spend $2$ or more hours per workday looking at their phones for personal activities."

  • Case Study 2 (Reviews.org Cell Phone Usage Habits): Surveyed N \n\approx 1,000 US adults ($18+$) in Q4 2025 via online Pollfish survey, weighted by age, gender, and census region ($ext{Margin of Error} = \pm 4\%$ at $95\%$ confidence level). Findings: Americans check phones $186$\,times/day; $84\%$ check within $10$\,minutes of waking; $56\%$ use phone during dinner; $68\%$ use on toilet; $40\%$ use on dates; $29\%$ use while driving; $72\%$ use at work; $41\%$ feel panic when battery drops below $20\%$; $53\%$ never go $> 24$\,hours without phone;$46\% consider themselves addicted.\n - *Case Study 3:* CDC BRFSS 2012 State-by-State Self-Reported Obesity Prevalence map.\n - **Association Claims:** Argue that two measured variables are correlated/linked.\n - *Validities Assessed:* Construct Validity, Statistical Conclusion Validity (strength, direction, significance), External Validity. *Internal Validity is not applicable.*\n - *Case Study (NPR Sports Defeat Diet):* Evaluated saturated fat and calorie intake among NFL fans following game outcomes. On Mondays following Sunday games, team defeat increased the saturated fat consumption index (1.16$) relative to victory ($0.91$) and elevated total calorie consumption ($404$\,kcal vs.$332\,kcal).\n - **Causal Claims:** Argue that one manipulated variable directly causes changes in an outcome.\n - *Validities Assessed:* **Internal Validity** (primary concern!), Construct Validity, Statistical Conclusion Validity, External Validity.\n - *Third-Variable Problem / Confound Examples:* Crime rates and ice cream sales correlate due to high summer temperature (third variable). Claiming psychiatric drugs cause mass shootings ignores underlying psychiatric conditions as a confound. Facebook use and divorce correlations (Clayton MU study 2013) cannot claim direct causality without experimental isolation.\n\n# Operationalization, Variables, and Statistical Variance\n\n- **Operational Definitions:**\n - The precise, concrete procedure used to manipulate or measure a theoretical variable in a research study.\n - For quantitative data, operationalizations must yield numerical values.\n- **Variable Classifications:**\n - **Categorical Variables:** Represent qualitative categories or kinds with no inherent numerical order (e.g., ice cream flavor, presence/absence of mental illness, gender, race, sexuality).\n - **Continuous Variables:** Represent numeric quantities, amounts, or ordered levels (e.g., height, weight, ice cream consumed in grams, age).\n- **Understanding Behavioral Variance:**\n - Psychological research investigates behavioral variability between people, across time, across situations, and in response to interventions.\n - **Classical Variance Equation:**\n    \text{Total Variance} = \text{Systematic Variance} + \text{Error Variance}\n - **Systematic Variance:** The proportion of total variance explained by the predictor variables included in the research model.\n - **Error Variance:** Variance attributable to unmeasured factors, individual differences, or measurement imprecise/poor reliability.\n- **Strength of Relationship / Effect Size Conventions (Cohen 1977 for \eta^2):**\n - Small Effect Size: 1\% of variance explained.\n - Medium Effect Size: 6\% of variance explained.\n - Large Effect Size: 15\% of variance explained.\n - *Real-World Benchmarks:* Ibuprofen for headache relief explains 2\%$of variance; Viagra for sexual dysfunction explains$14\% of variance.\n\n# Psychological Measurement, Reliability, and Construct Validity\n\n- **Classical Test Theory and Measurement Model:**\n  \text{Measured Score} = \text{True Score} + \text{Error}\n  \text{IQ Score} = \text{True IQ} + \text{Other Factors}\n- **Sources of Measurement Error:** Participant transient states (mood, fatigue, health, hunger), testing environment (time of day, room temp), and administrative factors (guessing, ambiguous wording).\n- **Strategies to Reduce Error:** Eliminate ambiguous phrasing, standardize administration training, and increase scale items.\n- **Reliability Definition & Relationship to Validity:**\n - Reliability refers to the consistency or stability of a measure.\n - A measure can be highly reliable without being valid, but a measure **cannot** be valid unless it is first reliable.\n- **Three Primary Methods for Assessing Reliability:**\n - **Test-Retest Reliability:** Same measure administered to same sample across two distinct time points. Measured via Pearson r$(requires$r > .80). Inappropriate for fluctuating trait states.\n - **Internal Consistency Reliability:** Degree to which individual scale items correlate at a single administration point. Measured via Cronbach's alpha (\alpha$; requires$\alpha > .80).\n - **Interrater Reliability:** Degree of agreement between independent raters coding behavior. Measured via Percent Agreement or Cohen's Kappa (\kappa).\n- **Subtypes of Construct Validity:**\n - **Face Validity:** Surface appearance that items assess the target construct.\n - **Content Validity:** Scale items cover the full theoretical domain defining the construct (e.g., Beck Depression Inventory assessing mood, sleep, appetite, suicidal ideation, and decision-making).\n - **Predictive Validity:** Scale scores accurately predict a future behavioral outcome (e.g., SAT scores predicting college GPA).\n - **Concurrent Validity:** Scale scores correlate with a criterion measured at the same time.\n - **Convergent Validity:** Scale scores correlate positively with existing measures of the same/related constructs (e.g., Need for Cognition correlating with Openness to Experience).\n - **Discriminant Validity:** Scale scores show zero or minimal correlation with theoretically distinct, unrelated constructs (e.g., Need for Cognition showing zero correlation with state anger or extraversion).\n\n# Survey Methodology, Question Wording, and Response Biases\n\n- **Response Biases & Biases:**\n - **Socially Desirable Responding:** Answering to present oneself in a favorable light.\n - **Acquiescence Bias ("Yea-Saying / Nay-Saying"):** Tendency to agree/disagree with all items regardless of content. Counteracted using **reverse-coded questions**.\n- **Flaws in Survey Question Wording:**\n - **Double-Barreled Questions:** Asking two distinct questions in a single item (e.g., "Should professors contact staff and administrators?").\n - **Loaded / Leading Questions:** Phrasing that biases respondents toward a specific answer.\n - **Negative & Double-Negative Wording:** Complex negative phrasing that impairs cognitive clarity.\n- **Framing & Word Choice Impact Examples:**\n - *2010 CBS News/NYT Military Survey:* Referencing "Gay Men & Lesbians" yielded significantly higher support for military service (58\%$favor) compared to referencing "Homosexuals" ($44\% favor).\n - *Loftus (1996) Vehicle Collision Study:* Changing the verb in speed questions ("smashed" vs. "hit" vs. "contacted") altered speed estimates (40.8$\,mph vs.$34.0$\,mph vs.$30.8$\,mph) and led$32\% of participants in the "smashed" condition to falsely recall broken glass one week later.\n- **Open-Ended vs. Closed-Ended Formats:**\n - *Open-Ended:* Unconstrained, rich qualitative data, free of researcher preconceptions, but difficult to code.\n - *Closed-Ended:* Standardized, easy to code, but options constrain output. Example discrepancy ("What is most important for children?"): Open format yielded 4.6\%$for "To think for themselves", whereas closed list format yielded$61.5\%$$.

• Response Order Effects: Primacy effect (selecting first option) and Recency effect (selecting last option). Diminished when options are presented visually.

Experimental Manipulation, Control, and Physiological Measures

• Key Experimental Design Features:

  • Longitudinal Design: Repeated measurements over extended time spans to establish temporal precedence.

  • Cover Story: A false narrative provided to participants to prevent demand characteristics.

  • Confederate: An actor working on behalf of the researcher who poses as a participant or bystander.

• Physiological Dependent Variables:

  • Galvanic Skin Response (GSR): Measures electrical skin conductance driven by autonomic nervous system arousal.

  • Electromyography (EMG): Measures micro-electrical activity produced by skeletal muscle movements (e.g., facial expressions).

  • EEG & fMRI: Measures brain activity (EEG tracks electrical signaling; fMRI tracks regional blood oxygenation and neural flow).

• Measurement Sensitivity Artifacts:

  • Ceiling Effect: Independent variable manipulation causes all scores to cluster at the maximum possible limit of the dependent scale.

  • Floor Effect: Independent variable manipulation causes all scores to cluster at the absolute minimum possible limit of the dependent scale.