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
Deception: Participants should not be misled about study aspects unless necessary to prevent demand characteristics.
Informed Consent: Participants must consent fully, knowing the study's aims.
Right to Withdraw: Participants have the autonomy to leave the study at any time.
Protection from Harm: Participants should be safeguarded from physical and psychological harm.
Confidentiality: Participant information must be kept private.
Privacy: Avoid intrusive questions about participants' private lives.
Debriefing: Participants need to learn the true aims of the study, particularly if deceived.
ANIMAL ETHICAL GUIDELINES
Numbers: Use only the minimum number of animals necessary for results.
Replacement: If possible, replace animals with computer simulations or video footage.
Pain and Distress: No infliction of suffering.
Reward, Deprivation, Aversive Stimuli: Animals should be rewarded for good behavior and not deprived or exposed to harmful stimuli.
Social Housing: Social species must not be isolated; non-social species must not be housed with others if harmful.
Caging: Avoid small, overcrowded cages.
Species and Strain: Incorrect species use can cause distress.
Anesthesia, Analgesia, and Euthanasia: Consider using anesthesia for procedures and euthanasia as a last resort only when no alternatives exist.
RESEARCH METHOD TECHNIQUES
EXPERIMENTS: Types include lab, field, and natural experiments.
OBSERVATIONS: Influencing factors include overt vs covert, participant vs. non-participant, structured vs. unstructured, and naturalistic vs. controlled.
SELF REPORT: Involves interviews (structured, unstructured, semi-structured) and questionnaires.
CASE STUDY: Focuses on an in-depth examination of one participant/group.
CORRELATION: Examines relationships between variables without causal implications.
LONGITUDINAL STUDIES: Following participants over time.
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.