Field, Survey, Correlational, and Experimental Control – Key Concepts

A comprehensive set of question-and-answer flashcards covering major concepts, definitions, designs, ethical issues, and control procedures from Chapters 6, 7, 8, 9, and 13 of the lecture notes.

What is naturalistic observation?

A low-constraint method in which researchers watch behavior in real-world settings without manipulating anything.

What is a case study?

An in-depth investigation of one person or a small group to understand a specific phenomenon in great detail.

Give two main reasons researchers use low-constraint methods such as naturalistic observation and case studies.

(1) To discover new ideas, behaviors, or hypotheses; (2) To study unique people or situations that cannot be easily manipulated in a lab.

List three key limitations of low-constraint research.

Little control over variables, low replicability, and inability to establish cause-and-effect relationships.

How do unobtrusive and participant observation differ?

Unobtrusive: observer remains outside the setting and is unnoticed; Participant: observer joins the group and records behavior from within.

What is archival research?

The analysis of existing records (e.g., hospital, government, or school data) to answer new research questions.

Why must field researchers be cautious about sampling?

Because they often cannot control who appears in the setting, findings should only be generalized to similar situations or populations.

Name two ethical requirements for covert field studies.

(1) Strong scientific justification; (2) Institutional Review Board (IRB) approval with safeguards for privacy and confidentiality.

Define the ex post facto fallacy.

Mistakenly inferring that A caused B simply because they occurred together in existing data.

State the two essential characteristics of low-constraint research.

Minimal control over variables and observation in natural, real-world settings.

What is measurement reactivity?

Participants alter their behavior because they know they are being observed.

What is experimenter reactivity?

Researcher expectations unintentionally influence participants’ behavior or data recording.

Why can naturalistic research negate but not establish a general proposition?

A single counter-example disproves an absolute claim, but many confirming cases cannot prove the claim for all possible situations.

Give one benefit of conducting experiments in field settings.

High external validity—findings are more likely to reflect real-world behavior.

Describe a nonequivalent control-group design.

A quasi-experiment comparing a treatment group with a non-randomly selected control group that does not receive the treatment.

Describe an interrupted time-series design.

Repeated measurements of one group before and after a naturally occurring event or intervention to detect its impact.

What is program evaluation?

The systematic assessment of whether a real-world program (e.g., educational, health) achieves its intended outcomes.

List two common problems faced during program evaluation.

Political or funding pressures and difficulty using random assignment or control groups.

Differentiate status surveys and survey research.

Status surveys describe current conditions; survey research tests relationships among variables measured with questionnaires.

Name four steps in designing a survey study.

Define information needed, create clear items, pre-test the questionnaire, and select a representative sample.

What is the main strength of survey research?

Ability to gather data from large, diverse samples quickly and inexpensively.

What is correlational research?

A non-experimental method that measures how strongly and in what direction two variables are related (without manipulation).

Explain why 'correlation is not causation.'

Because correlated variables may be linked by a third variable or by reverse causation; no manipulation or control confirms directionality.

What is differential research?

Comparison of two or more pre-existing groups (e.g., male vs. female) on a measured variable without manipulating group membership.

Define positive versus negative correlation.

Positive: variables increase or decrease together; Negative: as one variable increases, the other decreases.

Contrast cross-sectional and longitudinal designs.

Cross-sectional compares different groups at one time; longitudinal follows the same individuals across multiple time points.

What is a confounding variable?

An uncontrolled factor that varies with the independent variable and may account for observed effects.

What is an artifact in research?

A misleading result produced by the method or procedure rather than the variables of interest.

Explain the coefficient of determination (r²).

The square of a correlation coefficient, indicating the proportion of variance in one variable explained by the other.

What is a moderator variable?

A variable that changes the strength or direction of the relationship between two other variables.

Outline the five steps that turn an idea into a specific research hypothesis.

(1) Initial idea, (2) literature review, (3) precise problem statement, (4) operational definitions, (5) testable hypothesis.

Identify the three hypotheses encompassed by a typical causal research hypothesis.

Null hypothesis (no effect), confounding-variable hypotheses, and the causal hypothesis (IV causes DV).

List the four major types of validity.

Statistical, construct, internal, and external validity.

Give one threat to internal validity and its definition.

History: an external event during the study that could influence participants’ responses independently of the treatment.

Differentiate subject effects and experimenter effects.

Subject effects: participants change behavior due to expectations; Experimenter effects: researchers unintentionally bias outcomes.

What is statistical versus practical significance?

Statistical: result unlikely by chance (p < .05); Practical: effect size large enough to matter in real-world terms.

Why is random assignment considered powerful control?

It equalizes known and unknown participant characteristics across groups, reducing selection bias and confounds.

State the goal expressed by 'maximize experimental variance, control extraneous variance, minimize error variance.'

Enhance the true effect of the IV, prevent systematic influence of other variables, and reduce random noise for clear results.

Explain the logic of the F-ratio in ANOVA.

F = variance between groups (signal) divided by variance within groups (noise); a large F suggests the treatment affected means.

Describe a single-blind versus double-blind procedure.

Single-blind: participant or experimenter unaware of group assignment; Double-blind: both are unaware, reducing bias from expectations.

What is the balanced placebo design used for?

Separating pharmacological effects from expectancy effects by crossing actual substance intake with belief about intake.

Name the four population levels relevant to sampling.

General population, target population, accessible population, and sample.

Describe simple random, stratified random, and ad hoc sampling.

Simple random: every member equal chance; Stratified: random within subgroups; Ad hoc: convenient volunteers—limited generalization.

Give two common participant assignment methods other than free random.

Random-within-blocks assignment and matched random assignment.

Distinguish extraneous variance from error variance.

Extraneous variance comes from uncontrolled systematic factors; error variance is random, unsystematic variability within groups.

What is an ex post facto study and why is it weak?

A study examining existing groups after the fact without manipulation or pretest; lacks randomization and control for initial group differences.

How can naturalistic research enhance generalization compared with lab studies?

Behavior is observed in everyday contexts, so findings are more likely to apply across real-world settings.

Why can’t naturalistic research establish causality?

No manipulation of variables and many uncontrolled factors make alternative explanations impossible to rule out.

Summarize the key trade-off between laboratory and field research.

Lab offers control and causal clarity; field offers realism and external validity but less control.

Define statistical validity.

Accuracy of the numerical analyses and appropriateness of statistical tests used to draw conclusions.

What is regression to the mean?

Tendency for extreme scores to move closer to the average on subsequent measurements.

Give one technique for controlling subject and experimenter effects besides blinding.

Automation of procedures (e.g., computer-administered tasks) to minimize human influence.

Why is replication important, and what are its three main types?

Confirms reliability of findings; Types: exact, systematic, and conceptual replication.

Explain random within blocks assignment with an example.

Participants are grouped into blocks (e.g., by age); within each block one person is randomly assigned to each condition, balancing key variables.

What does the phrase 'covert observation requires strong justification' mean ethically?

Researchers must demonstrate that observing without consent is necessary, minimal risk, and IRB-approved to protect participant rights.

How does the balanced placebo design illustrate the difference between psychological and chemical effects?

By creating groups that receive or do not receive a substance and groups that think they did or did not, researchers separate expectancy from actual drug effects.

In survey research, what is a Likert-type item?

A statement rated on an agreement scale (e.g., strongly agree to strongly disagree) to measure attitudes or opinions.

What are two major difficulties unique to program evaluation?

Political pressures that may bias reporting and ethical concerns about withholding services from control groups.

Why is random sampling rarely achieved in human research?

Participation is voluntary and certain groups are harder to reach, preventing every population member from having an equal chance.

How can ethics constraints reduce external validity?

Restrictions on who can be included or how procedures are conducted may limit the diversity of participants and settings studied.

Give an example of holding a variable constant to control extraneous variance.

Testing only 18–22-year-old women to eliminate age and gender differences across groups.

What is diffusion of treatment, and why is it a validity threat?

When participants in different conditions share information, reducing differences between groups and contaminating results.