Psychological Research
Psychological Research: Chapter 2 Study Notes
The Foundation of Psychological Research
Research begins and ends with a question, such as: "How does television content impact children's behavior?" (FIGURE 2.1)
Historically, beliefs guided attempts to address mental illness; for example, trephination (making a hole in the skull) was believed to allow evil spirits to leave the body, thus curing disorders. (FIGURE 2.2)
Critical Thinking About Research Claims
Advertising campaigns often make claims based on "scientific evidence" that may actually be rooted in belief rather than fact.
When evaluating claims, it's essential to think critically by asking:
What is the expertise and credibility of the person making the claim?
What might the claimant gain if their claim is proven valid (e.g., financial, reputational benefit)?
Does the claim appear justified when weighed against the presented evidence?
What do other credible researchers and the broader scientific community think of the claim?
Notable Researchers in Psychology
Margaret Floy Washburn (1871–1939):
The first woman to earn a PhD in psychology.
Mary Whiton Calkins (1863–1930):
Conducted significant research on memory.
Established early experimental psychology labs in the U.S.
Opposed the behaviorist movement.
Francis Sumner (1895–1954):
The first African American to earn a PhD in psychology in .
Focused research interests on racial bias and educational justice.
Founder of Howard University's Department of Psychology.
Known as the "Father of Black Psychology."
Inez Beverly Prosser (1895–1934):
The first African American woman to earn a PhD in psychology.
Her research highlighted critical educational issues in segregated versus integrated schools.
Her work was highly influential in the landmark U.S. Supreme Court ruling, Brown v. Board of Education.
Social Science Research and the Scientific Method
Psychology is a science, and it utilizes the scientific method to understand behavior and mental processes.
The core of psychological research involves testing hypotheses that are derived from established theories.
Inductive vs. Deductive Reasoning
Psychological research employs both inductive and deductive reasoning. (FIGURE 2.4, FIGURE 2.5)
Deductive Reasoning (Theory Testing):
Moves from broad generalizations and existing theories to specific observations.
Example: All living things are made of cells () Humans are living things () Humans are made of cells ().
Inductive Reasoning (Theory Generation):
Moves from many specific observations to broader generalizations and the development of new theories.
Example: Observing several swans are white () Proposing all swans are white ().
The Scientific Method Explained
The scientific method involves a cyclical process: proposing hypotheses, conducting research, and then creating or modifying theories based on the results.
Scientists typically use inductive reasoning to formulate theories, which, in turn, generate specific hypotheses.
Theory:
A well-developed, comprehensive set of ideas that proposes an explanation for many observations.
Crucially, a theory must be falsifiable; meaning it must be capable of being disproven or shown to be incorrect.
Hypothesis:
A tentative, testable statement (a prediction) regarding the relationship between two or more variables.
It predicts how the world will behave if the underlying theory is correct.
Hypotheses are typically structured as "if-then" statements.
Falsifiability Examples
Falsifiability is a fundamental principle, asserting that for any hypothesis to be scientifically credible, it must be inherently disprovable.
Example from Freud's Theories (FIGURE 2.6):
Many specifics of Freud's theories, such as his division of the mind into the id, ego, and superego, have diminished in scientific favor because they are often not falsifiable.
If a theory posits that all behavior is unconscious, it becomes exceedingly difficult to falsify, as any behavior could be attributed to unconscious drives.
Contrast this with the hypothesis "all swans are white"; observing a single black swan can immediately falsify it.
Unfalsifiable Claim Example (Page 11):
A claim like "the Earth is younger than scientists state, having been created to appear older with deceptive fossils" is unfalsifiable.
Any evidence (fossils, geological data) presented against it can be dismissed by proponents as part of the initial deception, rendering it impossible to disprove scientifically.
Approaches to Research
Psychological research employs various methods:
Clinical or Case Studies
Naturalistic Observation
Surveys
Archival Research
Longitudinal and Cross-Sectional Research
Naturalistic Observation
Involves observing behavior in its natural setting.
Challenges:
Observer Bias: The researcher's expectations (consciously or unconsciously) can alter the results to fit expected outcomes. Jane Goodall's extensive naturalistic observation of chimpanzees is a classic example that requires careful management of observer bias. (FIGURE 2.8).
The act of observation itself can alter the behavior being observed (e.g., seeing a police car changes driving). (FIGURE 2.7)
Managing Observer Bias/Consistency:
Interrater Reliability: Assesses the consistency of observations made by different raters. High interrater reliability indicates that different observers agree on what they are seeing, reducing individual observer bias.
Clinical or Case Studies
Focuses intensively on one individual, typically someone in an extreme or unique psychological circumstance that distinguishes them from the general population.
Advantages:
Allows for deeply insightful and detailed information about the specific case.
Disadvantages:
Results are often difficult or impossible to generalize to the larger population due to the unique nature of the individual.
Example: The case of Genie, studied by psychologists after she was found at age having endured severe abuse and social isolation. Researchers were interested in the profound effect of social isolation on her development. (Page 15)
Surveys
A method for collecting data from a large number of people, often using questionnaires or interviews.
Can be administered in numerous ways, including electronically. (FIGURE 2.9)
Archival Research
Involves examining existing records or data sets.
These records can be archived as hardcopies or electronically. (FIGURE 2.10)
Longitudinal and Cross-Sectional Research
Longitudinal Research:
A research design in which the same group of individuals is studied repeatedly over an extended period.
Helps understand changes or developments over time.
Example: The Cancer Prevention Study- (CPS-) is a longitudinal study that helps us understand how smoking is associated with cancer and other diseases over decades. (FIGURE 2.11)
Cross-Sectional Research:
A research design where a researcher compares multiple segments of the population at the same moment in time.
Examples: Comparing dietary habits, physical activity, television viewing habits, or technology use across different age groups simultaneously. (Page 19)
Correlational Research
Correlation: Describes a relationship where two or more variables change together; as one variable changes, the other tends to change in a related way.
Correlation Coefficient ():
A numerical value ranging from to .
Indicates both the strength and direction of the relationship between variables.
The closer the absolute value of is to , the stronger the relationship.
Positive Correlation: Both variables change in the same direction, either both increasing or both decreasing.
Negative Correlation: The variables change in opposite directions; as one increases, the other decreases.
Scatterplots graphically display the strength and direction of correlations; points closer to a straight line indicate a stronger correlation. (Page 20)
Correlation Does Not Indicate Causation
A cause-and-effect relationship means that changes in one variable directly cause changes in another. This can only be determined through a carefully designed experimental research study, not solely through correlation.
Confounding Variable: An unanticipated external factor that influences both variables of interest, creating a false impression that one variable causes changes in the other.
Example: As ice cream sales increase, so does the overall rate of crime. While a strong positive correlation exists, ice cream sales do not cause crime. The confounding variable is temperature. As temperature increases, ice cream sales rise, and more people are outdoors, potentially leading to increased crime rates. (Page 21, FIGURE 2.13)
Further examples of spurious correlations can be found at:
https://tylervigen.com/spurious-correlations.
Illusory Correlations and Confirmation Bias
Illusory Correlations: Occurs when people perceive a relationship between two things when, in reality, no such relationship exists.
Example: Many people believe a full moon causes odd behavior, despite research showing no such relationship. (FIGURE 2.14)
Can contribute to the formation of prejudicial attitudes and discriminatory behavior.
Confirmation Bias: The tendency to seek out, interpret, favor, and recall information in a way that confirms one's preexisting beliefs or hypotheses, while ignoring evidence that disproves them.
Designing an Experiment
Defining and Measuring Variables
A hypothesis often stems from observations (e.g., watching violent TV leads to violent behavior). (FIGURE 2.15)
Operational Definition: A precise description of the specific actions and operations that will be used to measure the dependent variables and manipulate the independent variables in a study.
Experimental and Control Groups
Experimental Group: The participants who experience the manipulated independent variable; this group is designed to answer the research question.
Control Group: Participants who do not experience the manipulated independent variable; they serve as a crucial basis for comparison and help control for chance factors that might influence the study's results.
The experimental manipulation is the only designed difference between the experimental and control groups. Therefore, any observed differences between the two groups can be attributed to the manipulation, rather than chance or other factors.
In an experiment, manipulation of the independent variable is expected to cause changes in the dependent variable. (FIGURE 2.17)
Avoiding Bias and the Placebo Effect
Experimenter Bias: Occurs when a researcher's expectations or beliefs subtly (or overtly) alter the results of a study.
Participant Bias: Occurs when a participant's expectations or beliefs about the experiment influence their behavior, thereby altering the results.
Single-Blind Study: An experimental design where the researcher knows which participants are in the experimental group and which are in the control group, but the participants themselves do not. This controls for participant expectations and bias.
Double-Blind Study: The most rigorous design, where neither the researchers nor the participants know group assignments. This method controls for both participant and experimenter expectations and bias.
Placebo Effect: A phenomenon where people's expectations or beliefs influence or determine their experience in a given situation.
Example: If individuals believe a pill will improve their mood, their mood may genuinely improve simply due to this belief, even if the pill is inert (a placebo).
To determine if a medication has a genuine effect beyond the placebo effect, the experimental group receives the actual medication, while the control group receives a placebo treatment (e.g., a sugar pill). This setup is inherently a double-blind study. (FIGURE 2.16)
Any significant differences in outcomes between the groups can then be confidently attributed to the medication itself rather than expectations.
Selecting Participants
Participants (or subjects): The individuals involved in psychological research.
Population: The overall group of individuals that the researcher is interested in studying (e.g., all college students).
Populations are typically too large to include every member in a study.
Sample: A subset of individuals carefully selected from the larger population.
Random Sample: A subset of a larger population where every member of the population has an equal chance of being selected for participation.
This sampling method is highly preferred because it significantly increases the likelihood that the selected participants will be representative of the larger population across various demographics (e.g., sex, ethnicity, socioeconomic status). (FIGURE 2.18)
Assigning Participants to Groups: Random Assignment
Random Assignment: A fundamental method in experimental design where all participants have an equal chance of being assigned to either the experimental group or the control group.
Can be achieved using statistical software, random number generators, or simple methods like flipping a coin.
Crucially prevents systematic differences between the groups (e.g., in gender, age, prior experience).
Without random assignment, an experiment cannot establish a true cause-and-effect relationship, as any observed relationship could be due to preexisting differences among the groups.
Random assignment ensures that any significant differences found between the groups are highly likely to be the direct result of the experimental manipulation.
Issues to Consider: Manipulating Variables and Ethics
Manipulating Variables in Quasi-Experiments:
While random assignment is essential for establishing causation, some experimental designs are more complex.
Quasi-experimental: This type of experiment is conducted when an independent variable (such as sex, age, or a pre-existing condition) cannot be manipulated or when participants cannot be randomly assigned to groups (e.g., studying the effect of sex on spatial memory).
In quasi-experimental designs, a true cause-and-effect relationship cannot be definitively determined due to the lack of full experimental control.
Ethical Constraints in Research:
Some important research questions cannot be addressed using an experimental design because doing so would be unethical.
Example: Studying the effect of experiencing abuse as a child on levels of self-esteem. It is unethical to randomly assign participants to receive abuse.
Such questions must be investigated using alternative research approaches like case studies, surveys, or correlational designs, which do not involve manipulating potentially harmful variables.
Interpreting and Reporting Experimental Findings
Interpreting Experimental Data
After data collection, a statistical analysis is performed to determine the probability that any observed differences between experimental groups occurred purely by chance.
In psychology, results are typically considered "significant" if the probability that the differences occurred by chance is or less (p < .05).
A true experiment, which includes random assignment, manipulation of the independent variable, and control of extraneous variables, maximizes the likelihood that observed results are not due to chance.
Reporting and Replicating Findings
Research findings are usually disseminated through scientific journals, primarily aimed at an audience of professionals and scholars.
Peer-Reviewed Journal Articles:
Articles submitted for publication are rigorously read and evaluated by several other scientists (typically anonymously) who possess expertise in the subject matter.
These peer reviewers provide critical feedback regarding the quality, methodology, and significance of the manuscript before it is accepted for publication.
This process serves several vital functions:
It helps to filter out poorly conceived or poorly executed studies.
It often leads to suggested revisions that improve the clarity, rigor, and validity of the articles.
It ensures that the research methods are described with enough precision and detail to allow for replication by other researchers.
Replication:
The process of repeating a research study with different subjects or under different conditions to determine if the basic findings of the original study extend to other participants and circumstances.
Replication determines the reliability and generalizability of the original research design.
Replication studies can also incorporate additional measures that expand upon the original findings.
Ultimately, replication provides more evidence to support the original findings or, conversely, to cast doubt on those findings if they cannot be reproduced consistently.
Bad Science and Retraction: The Vaccine-Autism Myth
In a notable instance of flawed research, some peer-reviewed journals published studies claiming a link between routine childhood vaccines and the development of autism.
Subsequent large-scale, rigorous research unequivocally demonstrated that vaccinations are not responsible for causing autism.
Many of the original flawed studies were later retracted due to scientific misconduct, including the discovery that the lead researcher had a financial interest in establishing a link between vaccines and autism.
Unfortunately, the initial, incorrect claims received widespread publicity and continue to influence public perception, leading many people to still mistakenly believe that vaccinations cause autism.
Reliability and Validity
These are two crucial concepts for evaluating the quality of research measures and findings.
Reliability ():
The consistency and reproducibility of a given result.
Asks: "Would the same test or measurement tool yield the same results every time it is used under the same conditions?"
Also refers to whether the instruments and tools used to collect data do so in consistent, reproducible ways.
When a study involves observations by multiple people, it is critical that all observers make and record observations in the same consistent manner (interrater reliability).
Important Note: A reliable, consistent measurement does not necessarily mean that it is measuring something correctly or accurately. It can be consistently wrong.
Validity ():
The accuracy of a given result in measuring what it is designed to measure.
Asks: "Does a test or study truly measure what it is meant to measure?"
Crucial Relationship: A valid measure is always reliable, but a reliable measure is not always valid. (FIGURE 2.19, illustrating with target examples)
Ethics in Psychological Research
Research Involving Human Participants
Research involving human participants is subject to strict ethical guidelines.
Institutional Review Board (IRB):
A committee comprising administrators, scientists, and community members.
Exists at any research institution that receives federal support for research involving human participants.
The IRB regularly reviews all experimental proposals involving human participants and must approve a research proposal before it can proceed. (FIGURE 2.20)
Informed Consent: The process of fully informing a research participant about what to expect during an experiment and then obtaining their voluntary agreement to participate.
Includes detailing potential risks involved.
Explaining the implications and purpose of the research.
Notifying participants that their participation is entirely voluntary and they can withdraw at any time.
Assuring participants that any data collected will be kept confidential.
Deception
Sometimes, deception (purposely misleading experiment participants) is necessary to prevent participants' knowledge of the research question from affecting their behavior and thus the results.
Deception is only permissible if it is not considered harmful to the participants.
Debriefing: If an experiment involves deception, participants must be provided with complete and truthful information about the experiment's true nature and purpose at its conclusion.
The Tuskegee Syphilis Study (PHS Study) (FIGURE 2.21):
This infamous study (begun in ) highlighted the critical need for ethical guidelines.
Black men in the study who tested positive for syphilis were not informed of their diagnosis.
Although a cure (penicillin) was discovered in , it was intentionally withheld from the participants.
As a result, many participants unknowingly spread the disease, suffered severe health consequences, and many died.
Research Involving Animal Subjects
Research involving non-human animals is also subject to rigorous ethical oversight.
Institutional Animal Care and Use Committee (IACUC):
Similar to the IRB, this committee comprises administrators, scientists, veterinarians, and community members.
Reviews proposals for research involving non-human animals.
Justifications for Animal Research:
Approximately of psychological research involving animal subjects uses rodents or birds.
Animals are often used as substitutes for humans because many of their basic biological and behavioral processes are sufficiently similar to those in humans.
Animal models are used when the research questions would be unethical to explore with human participants.
Researchers are ethically obligated to aim to minimize pain or distress to animal subjects. (FIGURE 2.22)