Scientific Research Methods in Psychology

Research Goals and Descriptive Research Methods

• Four Primary Goals of Scientific Research:

  • Describe: Defining, observing, and documenting behaviors, attitudes, or mental processes as they naturally occur.
  • Predict: Identifying patterns, relationships, and associations between naturally occurring variables to anticipate future outcomes.
  • Explain: Determining cause-and-effect relationships by manipulating variables under controlled conditions.
  • Control: Applying scientific findings to influence, modify, or direct behavior and psychological outcomes predictably.

• Case Study Method:

  • Definition: An in-depth, exhaustive investigation of a single individual or a small, unique group.
  • Application: Utilized when investigating rare conditions, such as understanding the life, behavior, and psychological functions of a person with a rare brain disorder.
  • Methodological Rationale: Rare conditions present numerous unique variables that require multi-faceted analysis across many different aspects of the individual's life, which cannot be captured through standardized group studies.

• Self-Report Research Method:

  • Definition: A method relying on direct queries—such as surveys, questionnaires, or structured interviews—to gather data regarding participants' internal attitudes, feelings, or opinions.
  • Application: Used when direct observation is impossible, such as assessing gender attitudes across a large population like the city of Chicago.
  • Random Selection:
  • Necessary when studying large target populations (e.g., an entire municipality) to ensure every individual in the population has an equal probability of being chosen.
  • Produces a representative sample that accurately reflects the demography of the target population.
  • Response Rate and Sample Bias:
  • Researchers must critically evaluate both who receives the self-report instrument and who returns it.
  • Discrepancies in response rates jeopardize representativeness (e.g., mailing $1000$ surveys but receiving only $25$ back yields an unrepresentative sample that fails to accurately describe the broader city population).

Observational Research and Descriptive Scope

• Naturalistic Observation:

  • Definition: Systematic observation and recording of behaviors in their natural environment without researcher intervention or environmental manipulation.
  • Application Example: Describing and categorizing play patterns among children within a kindergarten classroom.
  • Observational Techniques:
  • Video Recording: Recording interactions on videotape and systematically reviewing and coding the footage later.
  • Two-Way Mirror: Observing participants behind a two-way mirror, allowing researchers to see the participants without being visible to them.
  • Direct Classroom Presence: The researcher physically sits inside the classroom. To prevent observer reactivity (e.g., children approaching, poking, or talking to the researcher), the observer must remain present for a couple of days prior to data collection so the children acclimate to their presence.

• Scope and Limitations of Descriptive Methods:

  • Scope: Case studies, self-reports, naturalistic observations, and laboratory observations yield detailed profiles that describe existing behaviors and attitudes.
  • Core Limitation: Descriptive research answers what behaviors or attitudes exist (e.g., describing gender attitudes in Chicago), but it cannot explain why individuals possess those specific attitudes or behaviors.

Correlational Research and Predictive Analysis

• Nature and Purpose of Correlational Design:

  • Predictive Function: Used to evaluate associations between variables in order to make predictions.
  • Measured Variables: Correlational designs analyze naturally occurring variables without experimental manipulation or forced conditions (e.g., measuring an individual's existing daily social media usage time alongside their measured level of depression, without inducing depression or mandating screen time).
  • Correlation Coefficient: Plotted statistical data undergoes analysis to produce a correlation coefficient, a numerical index quantifying the strength and direction of the relationship between variables.

• Three Essential Rules of Correlation:

  • Correlation Does Not Equal Causation: Demonstrating a statistical relationship between time spent on social media and depression level does not prove that social media usage causes depression, nor that depression causes social media usage.
  • The Third Variable Problem: An unmeasured extraneous variable may be driving the observed relationship, or the two measured variables may not be directly associated at all.
  • The Directionality Problem: Because ethical and practical constraints prevent assigning individuals to condition levels (e.g., researchers cannot randomly assign participants to be depressed or non-depressed), the direction of influence remains undetermined (whether social media usage affects depression or depression affects social media usage).

• Types and Framing of Correlations:

  • Positive Correlation: Variables move in the same direction simultaneously.
  • Example: As time spent on social media increases, measured depression increases.
  • Example: As the percentage of class attendance increases, overall GPA increases.
  • Negative Correlation: Variables move in opposite directions simultaneously.
  • Example: As time spent on social media increases, measured depression decreases.
  • Example: As the number of class absences increases, overall GPA decreases.
  • Analytical Rule: Determine whether a relationship is positive or negative strictly based on the specific verbal framing of the variables in the research statement rather than relying on intuitive assumptions.

Experimental Method and Testing Hypotheses

• Characteristics of Experimental Research:

  • Explanations and Causality: Experiments are the single research design capable of demonstrating cause-and-effect relationships (explanation).
  • Experimental Control: Achieved through rigorous control over environmental factors and systematically manipulating specific independent variables.

• Hypothesis Formulation:

  • Definition: A explicit, testable proposition predicting the relationship between variables.
  • Example: A psychologist predicts that a specific herbal remedy has a direct effect on an individual's mood.

• Experimental Structure and Variables:

  • Sample Size and Allocation: A total sample of $50$ participants is divided into two equal groups of $25$ participants.
  • Independent Variable (IV):
  • The factor manipulated or introduced by the researcher (e.g., the herbal remedy).
  • Methodological Clarification: Research participants themselves are never the independent variable (except in quasi-experimental designs where non-manipulable participant attributes, such as gender, define comparison groups).
  • Experimental Group (Group A): The treatment group that receives the active independent variable (the herbal remedy).
  • Control Group (Group B): The baseline comparison group that receives an inactive control substance.
  • Placebo:
  • An inactive substance that visually and physically resembles the active treatment (herbal remedy) but contains no active or beneficial ingredients.
  • Establishes a baseline control to evaluate true pharmacological or biological effects against participant expectations.

• Random Assignment:

  • Definition: Assigning participants to experimental or control conditions such that each participant has an equal chance of being placed in either group.
  • Methodological Function: Statistically minimizes the probability that extraneous individual differences (e.g., an uneven concentration of $25$ exceptionally happy or ecstatic individuals) end up systematically clustered in a single group.

• Dependent Variable (DV):

  • Definition: The variable measured by the researcher to evaluate the outcome or effect produced by the independent variable.
  • Example: Participant mood.
  • Final Experimental Analysis: Comparing the final mood measurements of Group A (experimental) directly against the mood measurements of Group B (control) to determine if the independent variable produced a statistically significant change.