Research Methods II Exam 2

Construct

unobservable psychological concept inferred from behavior (e.g., intelligence, memory) 🧠

Variable

anything measurable that can vary (e.g., reaction time, accuracy) 📊

Operational definition

exact way a variable is measured or manipulated 🔍

Independent variable (IV)

variable that is manipulated or predicts change ⚙️

Dependent variable (DV)

variable that is measured; outcome affected by IV 📈

Predictor variable

IV in correlational research 🔮

Criterion variable

DV in correlational research 🎯

Levels

different values of an IV (e.g., 0 mg vs 10 mg) 🔢

Subject variable

naturally occurring characteristic (age, sex) 👤

Manipulated variable

variable researcher controls (drug dose) 🧪

Control variable

factor kept constant to avoid influence 🛑

Confounding variable

variable that varies with IV and muddies results 🌪️

Reliability

consistency of a measure across time/situations 🔁

Test-retest reliability

same test given twice; correlate scores ⏳

Alternate forms reliability

two versions of same test; correlate scores 📄

Internal consistency

items on test measure same thing 🧩

Split-half reliability

correlate two halves of test ✂️

Interrater reliability

agreement between different scorers 👥

Intrarater reliability

consistency of same scorer over time 🎯

Validity

whether interpretation of results is appropriate ✔️

Construct validity

measure truly captures intended construct 🧠

Content validity

measure covers all parts of construct 📚

Convergent validity

correlates with similar measures 🔗

Discriminant validity

does NOT correlate with unrelated measures 🚫

Criterion-related validity

predicts relevant outcome 🎯

Predictive validity

predicts future outcome ⏭️

Concurrent validity

relates to outcome measured at same time 🕒

Face validity

appears to measure what it claims on surface 👀

Statistical conclusion validity

correct stats used to support conclusion 📊

Internal validity

ability to make causal claim 🔒

External validity

ability to generalize findings 🌎

Reliability vs validity

must be reliable to be valid, but not vice versa ⚖️

Frequency claim

describes rate of single variable 📊

Association claim

shows relationship between variables 🔗

Causal claim

states one variable causes change in another ⚡

Correlational design

measures variables without manipulating 📉

Bivariate design

examines two variables 🔢

Multivariate design

examines multiple variables 📊

Experiment

manipulates IV to test cause and effect 🧪

Covariation

IV and DV change together 📈

Temporal precedence

IV change happens before DV change ⏳

Third-variable problem

outside factor explains relationship ❓

Independent-groups design

different participants in each condition 👥

Within-groups design

same participants in all conditions 🔁

Random assignment

equal chance of being in any condition 🎲

Factorial design

study with 2+ IVs tested together ✖️

Cell

one condition in factorial design 📦

Main effect

overall effect of one IV 🧍

Interaction

effect of one IV depends on another 🔄

Crossover interaction

lines cross; opposite effects ❌

Spreading interaction

lines spread apart but don’t cross 📉

One-group posttest-only

treatment then measure; no comparison 🚫

One-group pretest-posttest

measure before and after treatment 🔄

Nonequivalent control group

has comparison group but no random assignment ⚠️

Interrupted time-series

multiple measures before and after treatment ⏱️

Control series design

time-series with control group ➕

Switching replications

treatment applied at different times to groups 🔁

A-B design

baseline then treatment (small-N) 🧍

Withdrawal (A-B-A) design

baseline, treatment, remove treatment ↩️

Multiple baseline design

compare behaviors/subjects without withdrawal 📊

Changing criterion design

gradually adjust performance goal 🎯

Alternating treatments design

rapidly switch treatments to compare ⚡

Longitudinal design

study same participants over time 📆

Cross-sectional design

compare different groups at one time 📍

Ethical behavior

acting morally, avoid harm, benefit society ✅

Historical rules

past bad studies → codes & reports 📜

Nuremburg Code

post-WWII, protect humans

Declaration of Helsinki

1964, ethical research guide

Belmont Report

1979, response to Tuskegee, laid principles

Beneficence

maximize benefits, minimize harm ⚖️

Autonomy

participants choose freely ✋

Justice

fair treatment for all 🤝

Beneficence & Nonmaleficence

do good, avoid harm 🌟

Fidelity & Responsibility

be professional, reliable 👩‍💼

Integrity

be honest, no cheating 📝

Justice

treat everyone equally 🤝

Respect for Rights & Dignity

protect privacy, voluntary participation 🔒

Primary risks

• physical harm 💪

• psychological harm 😔

• privacy breach 🔓

Informed consent

participants know purpose, procedures, risks/benefits, can refuse ✍️

Cannot consent

• minors 👶

• psychiatric patients 🧠

• cognitive impairment 🧩 (guardian approval needed)

Deception

allowed if safe & necessary 🤫 must debrief after

Debriefing

explain study fully, clear up deception, restore well-being 🧘

IRB

checks research for ethics 🏛️

Exempt

no risk (surveys)

Expedited

minimal risk (memory tests)

Full review

higher risk / vulnerable participants (kids, biological samples)

Fraud

fakes data → wastes resources, can harm people 💸⚠️

APA Ethics – 5 Principles

• Beneficence & Nonmaleficence

• Fidelity & Responsibility

• Integrity

• Justice

• Respect for Rights & Dignity