Approaches to Research Design
Approaches to Research Design
Introduction to Research Methods
Psychologists utilize diverse research methods to comprehend, describe, and explain behavior, as well as the underlying cognitive and biological processes.
Methods range from observational techniques to researcher-individual interactions, including simple questions, in-depth interviews, and controlled experiments.
Each method possesses unique strengths and weaknesses and is suited for specific research questions.
Observational studies: Yield vast information but have limited generalizability due to small sample sizes.
Survey research: Enables data collection from large samples, allowing for easier generalization, but collects limited information and is susceptible to self-report biases.
Archival research: Inexpensive data collection from existing records, providing insights, but researchers lack control over data collection methods or content.
Correlational vs. Causal Relationships:
Methods like observation, surveys, and archival research are inherently correlational, identifying relationships between two or more variables.
Correlational data cannot prove cause-and-effect relationships.
Establishing causation requires performing an experiment, which offers significant control over variables.
Experimental research, while powerful, often occurs in artificial settings, raising questions about the real-world validity of findings. Ethical concerns also prevent experimental pursuit of many psychological questions.
Clinical or Case Studies
Definition: Involves focusing in-depth on one person or a very small group of individuals.
Purpose: Scientists conduct case studies to gain extensive insight into rare or unique phenomena.
Example: Krista and Tatiana Hogan: Conjoined twins connected at the thalamus, a major sensory relay center.
Researchers are interested in the implications of this connection, such as one twin potentially experiencing the other's sensations.
These twins offer a vital resource for studying the brain, providing insight into shared sensory experiences, motor control, and the retention of distinct individuality.
Strengths:
Provides an unparalleled richness of information and a deep understanding of the studied individuals and phenomena.
Especially valuable when studying individuals with rare characteristics.
Weaknesses:
The primary benefit (in-depth focus on unique cases) is also its major drawback.
Limited Generalizability: Observations are difficult to apply to the broader population because the subjects are not typical.
Generalizing refers to the ability to extend research findings from a specific project to larger segments of society.
Naturalistic Observation
Definition: Observing behavior in its most natural context without intervention or manipulation.
Critical Aspect: The observer must be as unobtrusive and inconspicuous as possible to prevent subjects from altering their behavior due to awareness of being watched.
People tend to hide natural behaviors or perform social desirability when aware of observation (e.g., handwashing survey vs. actual observation).
Examples:
Preschool Playground Study (Fanger, Frankel, & Hazen, 2012): Researchers equipped children with wireless microphones and observed from a distance to study peer exclusion, leveraging subjects accustomed to observers.
Driving Behavior: A driver's behavior changes dramatically when followed by a police car compared to a deserted highway, demonstrating the impact of being observed.
Animal Studies: Jane Goodall spent nearly five decades observing chimpanzee behavior in Africa, providing invaluable insights into social hierarchies and communication. She faced criticism for giving chimps names, which some felt undermined objectivity.
Strengths:
High Validity/Accuracy: Data collected unobtrusively in a natural setting possesses high ecological validity, or realism.
Enhanced Generalizability: Findings are more applicable to real-world situations because behaviors are natural.
Avoids the issue of subjects modifying behavior if done correctly.
Weaknesses:
Difficult to Set Up and Control: Researchers have no control over when or if target behaviors occur.
Requires significant investments of time, money, and often luck.
Structured Observation: Sometimes used, where individuals are observed during specific, set tasks (e.g., Mary Ainsworth's Strange Situation to evaluate infant-caregiver attachment styles).
Observer Bias: Observers, often closely involved in research, may unconsciously skew observations to fit expectations.
Mitigation: Establish clear criteria for recording and classifying behaviors, and use inter-rater reliability (consistency of observations by multiple observers) to protect against bias.
Surveys
Definition: Questionnaires (paper-and-pencil, electronic, or verbal) used to gather data from research participants.
Strengths:
Efficient Data Collection: Surveys are typically quick to complete and easy to administer.
Large Samples: Allows researchers to collect data from substantially larger samples than other methods, improving generalizability.
A large, diverse sample better reflects the actual population diversity.
Weaknesses:
Limited Depth: Cannot collect the same in-depth information per person as case studies.
Inaccurate Responses: Participants may lie, misremember, or answer in a socially desirable way.
Measures of Central Tendency (from collected survey data):
: The most frequently occurring response.
: The middle value in a given data set.
: The arithmetic average of all data points. While useful for further analysis, the mean is highly sensitive to outliers.
Example: Attitudes toward Arab-Americans post-9/11 (Jenkins et al., 2012):
Researchers used surveys with direct and indirect questions to assess prejudice.
Participants reported no overt prejudice but showed less willingness for social interaction with Arab-Americans, suggesting subtle, underlying prejudice.
Archival Research
Definition: A research approach that uses existing records or data sets to answer research questions without direct interaction with participants.
Example: Examining academic records (completion time, course loads, grades, extracurriculars) to understand factors related to degree completion or risk factors for struggling students.
Strengths:
Cost-Effective: Considerably less time and money invested in data collection compared to methods involving direct participant interaction.
Access to large amounts of data.
Weaknesses:
No Control over Data Collection: Researchers cannot influence how or what information was originally gathered.
Research questions must be tailored to the existing data structure.
Potential for inconsistency between records from different sources, complicating comparisons.
Longitudinal and Cross-Sectional Research
Longitudinal Research
Definition: A research design where data is gathered repeatedly from the same group of individuals over an extended period.
Example: Surveying dietary habits of a group at ages , , and .
Strengths:
Powerful for observing how individuals change over time.
Reduces concerns about cohort differences (generational social/cultural experiences) influencing results, as the same individuals are tracked.
Effective for identifying predictive risk factors for diseases (e.g., Cancer Prevention Study-3 - CPS-3, which tracked thousands over decades to link smoking to cancer).
Findings from large-scale longitudinal studies can be generalized to the larger population with confidence.
Weaknesses:
Significant Investment: Requires incredible time and financial commitment from researchers and participants.
Results are not known for a considerable period (years, even decades).
High Attrition Rates: Participants may drop out due to moving, marriage/name changes, illness, death, or simply choosing to discontinue.
Researchers often recruit many participants, expecting substantial dropouts, and continuously check if the remaining sample still represents the larger population, making adjustments as needed.
Cross-Sectional Research
Definition: A research design that compares multiple segments of the population at the same time.
Example: Instead of tracking dietary habits for years, a researcher compares groups of -, -, and -year-olds simultaneously.
Strengths:
Requires a shorter-term investment compared to longitudinal studies.
Weaknesses:
Cohort Differences (Cohort Effects): Limited by differences existing between generations that are unrelated to age itself.
These differences reflect the unique social and cultural experiences of distinct generations, which can confound age-related findings.
Example: Support for Same-Sex Marriage: Younger people generally show more support than older individuals. Cross-sectional data cannot determine if this is due to aging making people less open, or if older individuals Hold different perspectives due to their distinct social climates during development.