research methods

VARIABLES AND CONTROLS

INDEPENDENT VARIABLE (IV)

  • Definition: A variable that is being manipulated to test for its impact on the dependent variable. It must have at least two conditions: an experimental condition and a control condition.
      - Experimental Condition: The group on which the IV is being tested, to observe if it affects participant behavior (e.g., the doodling group in Andrade).
      - Control Condition: The group where the IV is not applied, used as a comparison against the experimental condition to ensure any behavioral differences are due to the IV (e.g., the control condition in Andrade did not doodle).

  • Assignment of Participants: Participants can be allocated to conditions by the researcher’s choice or by random allocation.
      - Advantages: Reduces bias; increases validity by creating diverse participant types in each condition.
      - Disadvantages: Random assignment may result in individual differences (e.g., intelligence discrepancies between groups) that lower validity.

DEPENDENT VARIABLE (DV)

  • Definition: A variable that is measured to test if the IV influences it. Example: In Andrade’s study, the DV was the scores on monitoring and recall tasks.

CONFOUNDING VARIABLES

  • Definition: Also known as extraneous variables, these are any variables apart from the IV that may affect the DV. They can confound results if not controlled.
      - Example: In testing whether air conditioning (AC) affects concentration, factors like lighting could act as a confounding variable.

  • Types of Confounding Variables:
      - Participant Variables: Individual differences from participants (e.g., age, gender, intelligence).
      - Situational Variables: External factors from the environment affecting results (e.g., light, noise).
      - Uncontrollable Variables: Variables that cannot be managed or eliminated, which negatively influence validity.

CONTROLS

  • Definition: These are measures taken to reduce or eliminate confounding variables in a study.
      - Example: In studying AC’s effectiveness on concentration, one condition may have AC at a specific temperature (e.g., 16 degrees) while the control condition does not have AC.
      - Control Importance: Maintaining similar conditions like lighting across both groups ensures any noted differences in concentration stem from the IV alone.

  • Impact on Validity: High levels of control improve the reliability and validity of a study by standardizing procedures and reducing confounding variables.

VALIDITY AND ITS TYPES

VALIDITY

  • Definition: Refers to how accurately a study measures what it aims to measure.
      - Key for experiments: Validity increases as controls reduce the influence of confounding variables.

  • Demand Characteristics: If participants know the purpose of a study, this awareness may alter their natural behavior, thus lowering validity.

SOCIAL DESIRABILITY

  • Failure to provide honest answers in self-reports due to social norms can also reduce validity.

DOUBLE-BLIND TECHNIQUE

  • This method enhances validity by ensuring that neither participants nor observers know which condition participants are assigned to, thus avoiding demand characteristics and researcher bias.

ECOLOGICAL VALIDITY

  • Definition: The degree to which results can be generalized to real-life settings.
      - Field Experiments: High ecological validity when conducted in natural environments.
      - Lab Experiments: Typically low ecological validity due to artificial settings.

  • Mundane Realism: Related concept; the extent to which experimental tasks replicate real-life activities (e.g., helping versus electric shock scenarios in experiments).

TEMPORAL VALIDITY

  • Definition: The extent to which a measure remains relevant across different time periods.

CRITERION VALIDITY

  • The extent to which a measure correlates with a specific outcome, including two types:
      - Predictive Validity: Ability of a measure to forecast future outcomes (e.g., personality tests predicting job performance).
      - Concurrent Validity: Measurement correlation with established criteria taken simultaneously (e.g., depression scale correlating with clinical diagnosis).

RELIABILITY AND ITS TYPES

RELIABILITY

  • Definition: The consistency of results across repeated studies, which can be ensured via standardized controls.
      - A study should yield similar results if replicated under the same conditions.

INTER-RATER RELIABILITY

  • Involves consistency between two observers rating the same behavior. For example, if both observers rate aggression on a scale and agree, this indicates reliability.

INTER-OBSERVER RELIABILITY

  • Similar to inter-rater; focuses on observers reporting consistent behaviors rather than rating them (e.g., both observe a participant engaging in aggression).

TEST-RETEST RELIABILITY

  • Assessing the reliability of questionnaires or procedures by asking participants to engage with them multiple times (e.g., Yamamoto study on chimpanzee helping behavior).

SPLIT-HALF METHOD

  • A technique for checking questionnaire consistency by dividing it into two parts and comparing scores.
      - For instance, comparing responses to the first half versus the second half of a questionnaire.

GENERALIZABILITY

  • Definition: The degree to which study findings can be applied to larger populations.
      - Larger and more diverse samples increase generalizability.
      - For instance, a study sample of 10 females has low generalizability compared to a more diverse sample of 5000 people.

ORDER EFFECTS

DEFINITION

  • When the sequence of tasks impacts participants' behavior, potentially compromising validity.

PRACTICE EFFECTS

  • Improvement in performance due to repeated tasks (e.g., memorization of test answers from repetition).

FATIGUE EFFECTS

  • Decreased performance due to tiredness from task repetition.

RANDOMIZATION

  • A technique to minimize order effects by randomizing task order so that learned sequences do not occur.

COUNTERBALANCING

  • A method where participants complete two tasks (A and B) in varying sequences to control for order effects (e.g., half do A first, half do B first).

INDEPENDENT MEASURES DESIGN

  • Experimental design with different participants in each condition, mitigating practice and fatigue effects.

DATA

  • Definition: Results or findings from a study.
      - Quantitative Data: Numerical data allowing comparisons, indicative of outcome but lacking explanation.
      - Qualitative Data: Rich, detailed data representing behaviors and opinions, revealing the reasons behind actions.

ETHICS

DEFINITION

  • Ethical considerations are guidelines researchers must follow, separating categories for humans and animals.

HUMAN ETHICAL GUIDELINES

  1. Deception: Participants should not be misled about study aspects unless necessary to prevent demand characteristics.

  2. Informed Consent: Participants must consent fully, knowing the study's aims.

  3. Right to Withdraw: Participants have the autonomy to leave the study at any time.

  4. Protection from Harm: Participants should be safeguarded from physical and psychological harm.

  5. Confidentiality: Participant information must be kept private.

  6. Privacy: Avoid intrusive questions about participants' private lives.

  7. Debriefing: Participants need to learn the true aims of the study, particularly if deceived.

ANIMAL ETHICAL GUIDELINES

  1. Numbers: Use only the minimum number of animals necessary for results.

  2. Replacement: If possible, replace animals with computer simulations or video footage.

  3. Pain and Distress: No infliction of suffering.

  4. Reward, Deprivation, Aversive Stimuli: Animals should be rewarded for good behavior and not deprived or exposed to harmful stimuli.

  5. Social Housing: Social species must not be isolated; non-social species must not be housed with others if harmful.

  6. Caging: Avoid small, overcrowded cages.

  7. Species and Strain: Incorrect species use can cause distress.

  8. Anesthesia, Analgesia, and Euthanasia: Consider using anesthesia for procedures and euthanasia as a last resort only when no alternatives exist.

RESEARCH METHOD TECHNIQUES

  1. EXPERIMENTS: Types include lab, field, and natural experiments.

  2. OBSERVATIONS: Influencing factors include overt vs covert, participant vs. non-participant, structured vs. unstructured, and naturalistic vs. controlled.

  3. SELF REPORT: Involves interviews (structured, unstructured, semi-structured) and questionnaires.

  4. CASE STUDY: Focuses on an in-depth examination of one participant/group.

  5. CORRELATION: Examines relationships between variables without causal implications.

  6. LONGITUDINAL STUDIES: Following participants over time.

  7. CROSS-SECTIONAL STUDIES: Comparing different groups at a single point in time.

EXPERIMENTS

  • Definition of Experiments: A study manipulating IV to test its effects on DV (e.g., testing weather's impact on happiness).
      - Lab Experiment Strengths: High control and reliability, quantitative data, informed consent.
      - Weaknesses: Low ecological validity, potential demand characteristics.

FIELD EXPERIMENTS

  • Set in a natural setting, offering ecological validity and natural participant behavior.
      - Strengths: Real-world relevance, reduced demand characteristics.
      - Weaknesses: Less control and difficulty in replicating.

NATURAL EXPERIMENTS

  • Investigate naturally occurring IVs affecting DVs (e.g., gender differences in math performance).

OBSERVATION

  • Definition: Behavior analysis of participants which lacks IV or DV characteristics.
      - Covert vs. Overt: Awareness affects behavior observations.
      - Participant vs Non-Participant: The observer's involvement influences data accuracy.
      - Structured vs Unstructured: Use of checklists affects data richness.

SELF-REPORTS

  • Gathering data directly from participants through interviews and questionnaires, categorized as structured, unstructured, or semi-structured.

CASE STUDIES

  • Focus on detailed qualitative data from singular or small groups, allowing deep behavior insights, hindered by generalizability issues and potential researcher bias.

CORRELATION

  • Examines relationships without establishing cause-effect. The correlation coefficient measures strength and direction of relations.

SAMPLING METHODS

SAMPLING TECHNIQUES

  • Opportunity Sample: Convenience-based, leading to potential bias.

  • Volunteer Sample: Self-selecting, may result in skewed validity.

  • Random Sample: Equal selection chance, enhancing representativity but can limit specific characteristics.