Reading and Evaluating Scientific Research

Principles of Scientific Research

Objectivity in Scientific Research

Scientific research in psychology strives for objectivity, meaning that facts about the world can be observed and tested independently of the observer. However, subjectivity can influence research when prior beliefs, expectations, and experiences shape interpretation.

• Objectivity: Assumes facts are independent of the observer.

• Subjectivity: Knowledge is influenced by personal experiences and biases.

Characteristics of Quality Scientific Research

High-quality scientific research is defined by several key characteristics:

• Based on measurements that are objective, valid, and reliable

• Can be generalized

• Uses techniques that reduce bias

• Is made public

• Can be replicated

Scientific Measurement: Objectivity

Objective measurement ensures that the value of a quality or behavior is consistent regardless of who is measuring or the tool used. However, some margin of error is inevitable.

• Example: Your weight may differ slightly depending on the scale used (bathroom vs. gym), illustrating measurement error.

Why Do We Need Research?

• Facilitates communication

• Allows for public scrutiny and replication

What is a Research Question?

A research question guides the investigation and is based on:

• Common sense assumptions

• Observation in the real world

• Solving real-world problems

• Understanding how something works

Research Participants and Sampling

Who Are Research Participants?

• Study in samples: Select group of population

• Study in population: Entire group of people in interest

• Random sample: Technique in which every individual has an equal chance of being included

• Convenience samples: Those who are readily available, typically psychology students

Random Selection and Generalizability

Random selection is crucial for generalizability, which refers to the degree to which results can be applied to other populations or settings.

• Ensures sample accurately represents population

• Important for studies seeking generalizability (experiments)

• Example: Studying large groups for accuracy and reliability

Variables and Measurement

Variable vs. Operational Definition

Translating research questions into specific, testable procedures is essential.

• Variable: The object, concept, or event being controlled, manipulated, or measured

• Operational definition: Statements that describe the operations and specific measures used to observe variables

• Example: Measuring intoxication (variable) via blood alcohol level (operational definition)

Internal and External Validity

• Internal validity: How the study is conducted (control of variables)

• External validity: How applicable the findings are in real-world settings

Methods Used in Studies

• Naturalistic observation: Watching behavior in natural settings

• Self-report measures and surveys: Collecting data by asking participants to describe their own behaviors, attitudes, views, and perceptions

• Experimental designs: Manipulating variables to determine cause and effect

Evaluating Measures: Validity and Reliability

Validity

• Validity: The degree to which an instrument or procedure actually measures what it claims to measure

• High internal validity: High certainty that IV caused changes to DV

• Confound: A variable not of interest that varies along with the independent variable

Reliability

• Reliability: When a tool provides consistent and stable answers across multiple observations or points in time

• Test-retest reliability: Consistency of measure across test sessions

• Inter-rater reliability: Consistency across different raters

• Alternate-forms reliability: Consistency across different versions of the same test

Ecological Validity

• Results of a laboratory study can be applied to or repeated in the natural environment

Correlation and Experimental Methods

Correlation/Non-Experimental Method

Examines the strength of relationship between variables without manipulating them.

• Measures what is happening, does not manipulate the variable

• Correlation coefficient ranges from -1.0 to 1.0 (positive, negative, or zero)

• Higher value means stronger relationship

• Scatter plot illustrates the relationship

Examples of Correlation

• The relationship between texting speed & relationship drama

• Video games & aggression

• Time spent in traffic & happiness (negative)

• Number of tennis balls & dog (curvilinear relationship)

• Music coverage & Ivy immortality (nonlinear)

Correlation vs. Causation

• Correlation does not imply causation

• Third variable problem: A third variable may influence both variables, creating a false or misleading association

• Example: Kids with dogs may be happier, but it could be due to spending more time outside

Advantages and Disadvantages of Correlational Designs

• Advantages: Can establish relationships, good for describing behavior, can predict future behavior

• Disadvantages: Cannot infer causal direction, third-variable problem

Experimental Method

Research design that focuses on determining causal influences between variables.

• One variable is manipulated, and the other is measured or observed

• Random assignment of participants to experimental or control group

• Independent variable: What is being manipulated

• Dependent variable: What is being measured

• Control condition: Basis for comparison

• Example: Number of cups of coffee consumed (IV) and note-taking speed (DV)

Experimental Research: The Independent Variable

• Must have operationalization, at least two levels/conditions

• Ex: Treatment vs. placebo, High IV vs. Low IV

Experimental Bias

• Expectancy effect: Changes in participants' behavior caused by expectations of researcher/observer

• Double-blind designs: Prevent expectancy effects

• Demand characteristics/participant bias: Cues in research setting that lead participants to guess the purpose of the study and adjust their behavior

• Researcher bias: Bias in treatment of experiment, can question if it was for experimental reasons or their own bias

• Act of observation: When participants realize they are being observed, they may produce biased results (Hawthorne effect)

• Social desirability: Participants respond with what they think is expected

• Placebo effect: Measurable and experienced improvement in health or behavior that cannot be attributed to a medication or treatment

Ethical Guidelines for Human Research

• Must have informed consent

• Protect from harm and discomfort

• Deception and debrief (let them know what it is about after study has been done)

When is it Okay to Not Fully Inform?

• Research is purely observational

• Special populations

• Research purposes

• Required in cases where knowing the purpose would change their behavior

• Are not told the purpose of the study

• Misled given false purpose or not told

The Tuskegee Syphilis Study

• Ethical guidelines did not always exist

• In 1932, black men were recruited in an experiment studying syphilis

• Those tested positive were not informed they had a disease

• Although no cure existed at the time of study, a cure was found in 1947

Summary Table: Types of Validity and Reliability

Type	Definition	Example
Internal Validity	Degree to which the study controls for confounding variables	Random assignment in experiments
External Validity	Degree to which findings generalize to real-world settings	Naturalistic observation
Reliability	Consistency and stability of measurement	Test-retest reliability
Ecological Validity	Applicability of lab results to natural environments	Field studies

Key Equations

• Correlation coefficient: $r$ ranges from $-1.0$ to $1.0$

Additional info: Some explanations and examples have been expanded for clarity and completeness, including definitions of validity, reliability, and ethical guidelines.