AR

Research Methods in Design

Experimental vs. Non-experimental Methodologies

  • Experimental Methodology:

    • A systematic approach carried out under controlled conditions. Example: A study testing the effect of a new drug on depression, where participants are randomly assigned to receive the drug or a placebo.

    • Aims to test a hypothesis. Example: Testing whether caffeine improves reaction time.

    • Establishes a causal relationship between the independent and dependent variables. Example: Showing that the drug directly causes a reduction in depression symptoms.

    • Explains behaviors.

  • Non-experimental Methodologies:

    • Used when a controlled experiment is not possible or ethical. Example: Studying the effects of a natural disaster on mental health.

    • Describes behaviors but cannot explain them. Example: Observing that individuals exposed to a natural disaster experience higher rates of anxiety, without determining why.

    • Cannot establish a causal relationship between variables. Example: Observing a correlation between income and education level, but not proving that higher income causes higher education.

    • Includes case studies, correlational studies, meta-analysis, and naturalistic observation.

Non-Experimental Methods

  • Case Study:

    • Examines an individual, group, event, or situation. Example: Studying a patient with a rare psychological disorder in detail.

    • Provides detailed information and insight.

    • Risk of the Hawthorne effect: subjects alter their behavior when aware of being observed. Example: Factory workers increasing productivity when they know they are part of a study

  • Correlational Studies:

    • Gain insight into the relationship between two variables. Example: Examining the relationship between exercise and happiness.

    • Determines the strength of the relationship.

    • Correlation does not equal causation.

    • Risk of the third variable problem: an outside variable impacts the study that was not accounted for. Example: Ice cream sales and crime rates both increase during summer, but neither directly causes the other

  • Meta-Analysis:

    • A statistical technique.

    • Combines results of multiple studies on the same topic to reach a conclusion. Example: Combining multiple studies on the effectiveness of cognitive-behavioral therapy for anxiety.

    • Studies studies instead of participants.

  • Naturalistic Observations:

    • Researchers observe individuals in a real-world setting. Example: Observing children's behavior on a playground

    • Aims to gather authentic data.

    • Issue: Observers may lack proper context depending on the length of observations.

Designing a Study

  • Hypothesis:

    • A specific, testable prediction about the relationship between two or more variables. Example: "Increased screen time is associated with lower academic performance in high school students."

    • Must be falsifiable (can be proven wrong).

    • Example: "Students who use my ultimate review packet will score higher on the AP Psychology exam compared to students who do not use the packet."

  • Theory:

    • Supported by data from research.

    • Explains a question, thought, or phenomenon. Example: The theory of cognitive dissonance explains why people experience discomfort when holding conflicting beliefs

    • Based on tested hypotheses.

    • Allows prediction of future outcomes.

  • Operational Definitions:

    • Outline the exact procedures used in a study.

    • Define how variables are measured or manipulated.

    • Allows for replication of the study under the same conditions.

    • Example: Studying effects of sleep on academic performance.

      • Hypothesis: "Students who get more sleep the night before the exam will score higher on the exam compared to students who get less sleep."

        • More sleep: \geq 8 hours of continuous sleep tracked with Apple Watches.

        • Less sleep: < 8 hours of continuous sleep.

        • Exam: AP Psychology National exam, performance measured on a scale of 1 to 5.

Variables

  • Independent Variable (IV):

    • Manipulated or controlled by the researcher. Example: The dosage of a drug given to different groups of participants

    • The cause.

  • Dependent Variable (DV):

    • The outcome being measured. Example: The severity of depression symptoms in each group.

    • The effect.

    • Example: Ultimate review packet (IV), Exam score (DV).

  • Confounding Variables:

    • Factors other than the IV that could impact the DV. Example: Pre-existing mental health conditions

    • Variables the researcher couldn't remove.

    • Example (Sleep Study): Study habits, stress, overall health.

  • Control:

    • More control in a study = fewer confounding variables.

    • Trying to control an experiment too much may lead to an inauthentic environment, creating new confounding variables.

Participants

  • Population:

    • The entire group being studied. Example: All college students in the United States

  • Sample:

    • Selected individuals from the population to represent the whole. Example: A group of 100 students from various colleges across the US.

    • Example: Student body (population), selected students (sample).

  • Random Sampling:

    • Each individual in a population has an equal chance of participating. Example: Using a random number generator to select participants from a list.

  • Stratified Sampling:

    • Population divided into subcategories. Example: Dividing a population into age groups and randomly sampling from each group.

    • A random sample taken from each subcategory.

  • Representative Sample:

    • The sample group represents all people in the population. Example: A sample that includes participants of different ages, genders, and ethnicities in proportion to the general population.

  • Sampling Bias:

    • The sample does not accurately represent the population. Example: Only surveying students from one specific college.

    • Occurs when the selection process is flawed.

    • Example: Convenience sampling (selecting individuals based on availability).

  • Generalizability:

    • Extent to which study findings can be applied to the larger population. Example: If a study on test anxiety is conducted with a representative sample, the results can be generalized to the larger population of students.

Experimental and Control Groups

  • Experimental Group:

    • Receives the independent variable. Example: A group receiving a new medication.

  • Control Group:

    • Receives a placebo. Example: A group receiving a sugar pill instead of the actual medication.

    • Placebo: Something close to the IV but missing a key component.

  • Random Assignment:

    • Participants are randomly assigned to the control or experimental group. Example: Using a coin flip to decide which group each participant is assigned to.

  • Appropriate Representation:

    • A sample accurately reflects the population's demographics. Example: Ensuring that a study on heart disease includes participants of different ages, genders, and ethnicities in proportion to the general population.

    • Increases validity and generalizability.

    • Reduces bias.

Quasi-Experiment

  • Used when random assignment is not ethical or possible. Example: Studying the effects of a new teaching method on existing classrooms without reassigning students.

  • Does not determine cause and effect because group differences are not controlled by random assignment.

  • Lacks random assignment of participants.

  • Experimental methods must use random assignment and will always involve independent and dependent variables

  • Non-experimental methods will not always include random assignment

Procedures

  • Single-Blind Procedure:

    • Participants do not know if they are in the experimental or control group. Example: In a drug trial, participants do not know if they are receiving the actual drug or a placebo.

    • Prevents social desirability bias and placebo effect.

      • Social desirability bias: Participants skew answers to create a favorable impression. Example: Over reporting positive behaviors to appear favorable

      • Placebo effect: An individual's physical or mental state improves after taking a placebo because they believe they are taking the real substance.

  • Double-Blind Procedure:

    • Neither participants nor researchers know who is in each group. Example: Researchers administering the drug do not know which participants are receiving the actual drug and which are receiving a placebo.

    • Counters experimenter bias and social desirability bias.

      • Experimenter bias: The researchers' expectations, preferences, or beliefs influence the outcome of the study unknowingly. Example: A researcher subtly influencing participants to respond in a way that confirms their hypothesis.

Measurements

  • Qualitative Measures:

    • Collect non-numerical data. Example: Interview transcripts

    • Provide detailed, descriptive insights into participants’ thoughts, feelings, and behaviors.

    • Example: Structured interviews with open-ended questions.

    • Produce descriptive and subjective information that is hard to replicate but provide insight into participant's experiences.

  • Quantitative Measures:

    • Collect numerical data. Example: Test scores

    • Analyzed statistically to identify relationships, patterns, and differences.

    • Example: Likert scale (rating agreement with statements on a scale).

    • Produce objective information that measures variables in numerical form allowing statistical analysis and replication.

Protecting Participants

  • Informed Consent:

    • Participants understand the necessary information to make an informed decision. Example: Explaining the study's purpose, procedures, risks, and benefits to participants before they agree to participate.

    • Aware of the risks of the study.

    • Free to choose whether to participate.

  • Informed Assent:

    • Participant is legally unable to provide full consent (e.g., a minor). Example: Obtaining agreement from a child to participate in a study, along with parental consent.

    • The participant must agree to the study along with a parent or guardian.

  • Ethical Studies:

    • Create a positive environment for subjects.

    • Participants trust the researcher.

    • The study has a net benefit for society.

    • Integrity and transparency with participants.

      • Debriefing: Explaining information about the study at the end. Example: Informing participants of any deception used in the study and the true purpose of the study.

  • American Psychological Association (APA):

    • Established in 1892 as a governing board to study behavior.

    • Created the first ethics committee in 1947 to create standards for all psychological research.

  • Institutional Review Board (IRB):

    • Created in 1974 to protect human participants.

    • All colleges and universities use the IRB to conduct any experiments or research studies in psychology.

    • Looks at proposed research studies with human participants.

    • Will reject a study if participants are not being protected.

  • Institutional Animal Care and Use Committee (IACUC):

    • Regulates and oversees animal care and research, teaching, and testing with animals.

  • APA Ethical Standards:

    • Must be followed by all researchers to protect human and animal subjects.

    • Respect people's rights and dignity.

Conclusions: Peer Review and Replication

  • Peer Review:

    • Experts in the field assess the study's methodology, data, and conclusions before publication. Example: Submitting a research paper to a journal, where it is reviewed by other researchers in the field before acceptance.

  • Replication:

    • Other individuals conduct the study again to check the original findings and verify the results. Example: A different research team conducting the same experiment to see if they obtain similar results.

  • Peer review and repeated replication allow scientific research to evolve and maintain high standards.