2.2

Page 1

Learning Objectives

• By the end of this section, you will be able to:
  • Describe the different research methods used by psychologists
  • Discuss the strengths and weaknesses of case studies, naturalistic observation, surveys, and archival research
  • Compare longitudinal and cross-sectional approaches to research
  • Compare and contrast correlation and causation

Overview of research methods in psychology

• Psychologists use a variety of methods to understand, describe, and explain behavior and the cognitive and biological processes underlying it.
• Methods can be observational (no direct manipulation or interaction) or involve direct researcher-participant interaction (from simple questions to in-depth interviews) or well-controlled experiments.
• Each method has unique strengths and weaknesses and is suitable for different research questions.
• Examples of trade-offs:
  • Observation: produces rich, detailed information but small sample sizes limit generalizability to the larger population.
  • Surveys: easy to collect data from large samples, aiding generalizability, but information per participant is often limited and self-report data can be biased or inaccurate.
  • Archival research: inexpensive and leverages existing records, but researchers have no control over how data were collected or what data exist.
• All the methods described thus far are correlational in nature. They can identify relationships between variables but cannot establish causation.
• Correlational data can reveal relationships, but to claim causation you must conduct an experiment.
• Experimental research offers high control over variables but can be conducted in artificial settings, which may threaten real-world applicability (ecological validity).
• Some questions cannot be pursued with experiments due to ethical concerns.

Key concepts to remember

• Correlation does not imply causation. Experimental control is required to infer cause-and-effect.
• External validity (generalizability) vs. internal validity (control of confounding variables).
• Ethical considerations can constrain the kinds of questions that can be pursued experimentally.

Notes on research design trade-offs

• Observational and correlational methods are valuable for exploring questions and generating hypotheses, but their findings are inherently limited in establishing causal direction.
• Experiments provide strong evidence for causation but may sacrifice everyday realism.
• Researchers often triangulate findings using multiple methods to build a robust understanding of a phenomenon.

Mathematical reminder: central tendency (for survey data)

• The central tendency of a data set can be described by:
  • Mode: the most frequently occurring value
  • Median: the middle value when data are ordered
  • Mean: the arithmetic average, which is especially informative for further statistical analyses
• The mean is defined as
  $\text{mean} = \frac{1}{n} \sum<em>{i=1}^{n} x</em>i$
• Note: the mean is sensitive to outliers, which can distort interpretation of typical values.

Summary of Page 1

• Psychology uses multiple research methods, each with strengths and weaknesses.
• All discussed methods are correlational unless experiments are conducted.
• Experimental research offers control and potential causal inferences but may reduce realism and raise ethical constraints.
• Surveys and archival research offer breadth and efficiency but come with limitations in depth or data quality/control.

Page 2

Clinical/Case Studies in psychology

• Case studies focus in-depth on a single person or a very small number of individuals (often 10–20).
• The richness of information in case studies can provide deep insights into a phenomenon that may be rare or unique.
• Krista and Tatiana Hogan (conjoined twins) provide an example of a case study with unique neurological features:
  • They are connected at the head, with evidence of a connection in the thalamus, a major sensory relay center.
  • The thalamus is a gateway for sensory information to reach higher brain regions; hence, a shared thalamic connection could theoretically allow one twin to experience sensations of the other.
  • This rare condition offers a valuable resource for studying brain processing, though generalizability to the broader population is limited.
  • Over time, Krista and Tatiana have shown that they remain two distinct individuals, illustrating both shared and independent aspects of perception and motor control.
• Pros of case studies:
  • Rich, in-depth information.
  • Potential to study rare phenomena that would be difficult to recruit for in larger samples.
• Cons of case studies:
  • Limited generalizability to the larger population due to unique characteristics of the individuals studied.
  • Difficulty in applying findings to explain behavior in the average person.
• Generalization refers to applying findings from a case study to larger groups; this is often challenging with rare or unusual cases.

Why use case studies?

• When researchers encounter rare or extreme cases that yield valuable hypotheses or theory-building insights.
• To obtain a detailed, contextual understanding of a phenomenon that might not be captured by other methods.

Notes from the Hogan case in context

• The Hogan twins illustrate how a single case can raise questions about brain connectivity, sensory integration, and the potential for shared experiences.
• Ethical and practical considerations include ongoing family consent and the long-term follow-up that rare cases may require.
• Case studies may inspire laboratory or cross-sectional research to test the generalizability of observations.

Summary of Page 2

• Case studies provide unparalleled depth for rare phenomena but have limited generalizability.
• Real-world example (Krista and Tatiana Hogan) highlights how unique biological features can inform brain research, while also illustrating limitations in extrapolating to broader populations.

Page 3

Naturalistic Observation: key concepts and methods

• Naturalistic observation involves watching behavior in its natural context without interference.
• A major challenge is that people (or animals) may alter their behavior if they know they are being watched, threatening ecological validity.
• To obtain more accurate information, researchers often use inconspicuous observation:
  • Example: observers blend into the environment (e.g., pretending to do something ordinary while quietly recording behavior).
  • In a classroom hand-washing example, a covert observer might stand by a sink, blending in with the setting to observe true behavior.
• Suzanne Fanger and colleagues studied preschool peer exclusion by observing on a playground. They used wireless microphones and a laboratory preschool setting to minimize observer effects and maintain natural behavior.
• In naturalistic observation, it is critical that the observer remains unobtrusive because awareness of being observed changes behavior.

Naturalistic observation in animals

• The same approach applies to studying animals in their natural habitats to understand social structures and communication.
• Jane Goodall's decades-long study of chimpanzees in Africa is a famous example. She observed chimpanzee behavior in the wild for nearly five decades.
• Ethical and methodological considerations included whether naming individual animals might compromise objectivity or psychological detachment. Critics argued that personalizing animals (e.g., giving them names) could affect the perceived objectivity of the study.

Strengths and limitations of naturalistic observation

• Greatest strength: high ecological validity because behavior is observed in natural settings without artificial manipulation.
• Strong external validity when findings generalize to real-world contexts.
• Greatest limitation: lack of experimental control, which can make it difficult to determine causal relationships or to replicate precisely.
• Other practical challenges: time, money, and luck; researchers may face long durations before observations of interest occur.
• Naturalistic observation can involve structured observation, where researchers observe behavior during specific tasks or phases.

Key terms

• Ecological validity: the extent to which findings generalize to real-world settings.
• Observer bias: when observers' expectations influence how they record or interpret behaviors.
• Inter-observer reliability: the degree to which different observers record and classify behavior consistently; often used to assess reliability.

Summary of Page 3

• Naturalistic observation provides rich natural data but has limited experimental control.
• Inconspicuous observation helps preserve natural behavior, though it raises ethical and methodological considerations.
• The technique is applicable to both humans and animals, with famous exemplars like Jane Goodall.
• Structured observation and inter-observer reliability are tools to improve rigor in observational research.

Page 4

Naturalistic observation continued; ethical and practical concerns

• The principle of inconspicuous observation is violated in settings like reality TV where participants are followed by cameras and interviewed for confessionals; this compromises the authenticity of behavior and reduces naturalness.
• Benefits of naturalistic observation include high validity of data collected unobtrusively in natural settings, leading to greater ecological validity and generalizability to real-world contexts.
• Downsides include difficulty in controlling when observations occur and the potential for missing data if nothing occurs to observe for long periods.
• Animal observation in the wild requires remaining unseen or unobtrusive to avoid altering natural behaviors of the species under study.
• The balance between ecological validity and experimental control is a core consideration in choosing observational methods.

Structured observation and the Strange Situation

• Some observational studies use a structured protocol in which specific tasks or phases are used to elicit observable behaviors.
• An iconic example is the Strange Situation by Mary Ainsworth, used to evaluate infant attachment styles:
  • Infants and caregivers enter a room with toys; the procedure includes several phases such as a stranger entering, the caregiver leaving, and the caregiver returning.
  • The infant’s behavior upon reunion with the caregiver is particularly informative for classifying attachment style.
• Other potential issues in observational research include observer bias, which can be mitigated by:
  • Establishing clear criteria for what constitutes specific behaviors.
  • Having multiple observers record the same events to test inter-rater reliability.
• Inter-rater reliability is a measure of the consistency of observations between different observers and helps ensure data accuracy.

Summary of Page 4

• Naturalistic observation emphasizes unobtrusiveness and ecological validity but requires careful handling of observer bias and reliability.
• Structured observation (e.g., Strange Situation) provides more standardized data on specific behaviors.
• Reliability checks, including using multiple observers, strengthen the credibility of observational data.

Page 5

Surveys: collecting data from large samples

• Surveys are a practical method for gathering data from large samples via questionnaires, either on paper, electronically, or verbally.
• Advantages:
  • Quick to administer and scalable to large populations.
  • Facilitate generalization when the sample is large and diverse.
• Disadvantages:
  • Depth of information per individual is limited compared to case studies.
  • Self-report data are vulnerable to bias: social desirability, misremembering, or intentional misreporting (lying to appear favorable).
• Central tendency and descriptive statistics are often used to summarize survey results. The three common measures are:
  • Mode: most frequently occurring response
  • Median: middle value in an ordered data set
  • Mean: arithmetic average
  • The mean is particularly useful for subsequent analyses but is sensitive to outliers:
    $\text{mean} = \frac{1}{n} \sum<em>{i=1}^{n} x</em>i$
• Strengths vs. weaknesses of surveys relative to case studies:
  • Surveys sample more people and better reflect population diversity, aiding generalizability.
  • Case studies provide depth and context, but generalizing from a single or few cases is challenging.
  • Surveys trade depth for breadth; responses may not accurately represent true attitudes or behaviors.

Applied example: attitudes toward Arab-Americans after 9/11

• Jenkins, Ruppel, Kizer, Yehl, and Griffin (2012) conducted a survey with 140 participants using 10 questions.
  • Direct questions assessed overt prejudicial attitudes toward various ethnic groups.
  • Indirect questions evaluated willingness to interact with people of Arab-American descent in different settings.
  • Findings suggested participants reported low overt prejudice, but responses indicated reduced social interaction willingness with Arab-Americans, implying subtle forms of prejudice not captured by direct questions.

Archival research: using existing records

• Archival research relies on existing records or data sets without direct participant interaction.
• Advantages:
  • Less time and money required for data collection.
  • Useful for answering questions when experimental data are unavailable.
• Limitations:
  • Researchers have no control over how data were collected or what data exist.
  • Inconsistencies across data sources can complicate comparisons and synthesis.
• Example: archiving educational records to study time-to-degree, course loads, grades, and extracurricular involvement to identify risk factors for academic struggles (Figure 2.10).
• In comparing archival research to other methods, researchers must tailor questions to the structure and limitations of existing data.

Summary of Page 5

• Surveys offer breadth and generalizability but can sacrifice depth and are susceptible to self-report biases.
• Archival research is efficient but limited by data quality and consistency; there is no direct interaction with participants.

Page 6

Longitudinal and Cross-Sectional Research: studying change over time

• Researchers sometimes want to examine how people change over time, such as during development or aging.
• Longitudinal research:
  • Data are gathered from the same individuals repeatedly over an extended period.
  • Example: dietary habits surveyed at age 20, retested at age 30, then at age 40.
  • Advantages: controls for cohort effects by following the same individuals, enabling clearer interpretation of changes over time.
  • Disadvantages and limitations:
  • Very time-consuming and expensive.
  • High dropout (attrition) rates; participants may move, change names, become ill, or die; even in the absence of life changes, some may discontinue participation.
  • Researchers recruit many participants to anticipate dropouts and continually assess whether the sample remains representative of the population.
• Cross-sectional research:
  • Researchers compare different segments of the population at a single point in time.
  • Example: compare dietary habits of 20-, 30-, and 40-year-olds at one time point rather than following them over decades.
  • Advantages: faster and less costly than longitudinal studies.
  • Limitations: cohort effects—differences between generations (social, cultural experiences) that may be mistaken for aging effects.
  • The key challenge is disentangling age effects from cohort effects.

Why longitudinal studies are powerful

• Useful for investigating disease risk factors and long-term outcomes in large samples.
• CPS-3 (Cancer Prevention Study-3) is an ongoing American Cancer Society longitudinal study tracking hundreds of thousands of participants over 20 years.
  • Initial data: participants complete a life-history survey, including family history and risk factors.
  • Follow-up: periodic surveys over time to determine who develops cancer and who does not.
  • Historical significance: early longitudinal studies helped establish links between smoking and cancer.
  • CPS-3 illustrates the scale and potential of longitudinal designs for identifying predictive risk factors.

Figure references

• Figure 2.11 (described): Longitudinal research like CPS-3 helps us understand how smoking is associated with cancer and other diseases.

Limitations of longitudinal research

• Requires a substantial time commitment from researchers and participants.
• Financially demanding due to long durations and large samples.
• Attrition reduces sample size over time, potentially biasing results if dropouts are not random.
• Researchers must continually assess whether the remaining sample still represents the larger population and adjust as needed.

Summary of Page 6

• Longitudinal designs provide insight into changes over time and help establish temporal sequences, but they are resource-intensive and susceptible to attrition.
• Cross-sectional designs are efficient but vulnerable to cohort-confounding effects; careful interpretation is required.