Study Notes - Experimental Statistics and Its Applications in Psychology

Definition: Experimental statistics is a branch of statistics that involves using statistical methods to test hypotheses and analyze data collected from experiments.
Objective: Assess if control groups show significant changes compared to experimental groups.
Context: Commonly applied in psychology where large sample sizes are often not feasible, necessitating observational studies.

Challenges: Limited sample sizes in psychological studies complicate the execution of experimental statistics.
Observational Studies: Frequently conducted due to the difficulty in testing the entire population.
Impact of Small Samples: A small sample can provide insights on whether manipulations (e.g., vitamin C dosage) yield observable effects.
- Example: Uncertainty around beneficial doses of vitamin C.

Ongoing Research: Nutritional science continuously evolves as new research updates recommendations (e.g., vitamin C supplementation).
Complexity of Factors: Individual reactions to supplements may vary due to numerous unknown factors.
- Example: Garlic supplementation and its purported effect on cholesterol and glucose levels.

Experiential Science: Unlike physics that can use precise mathematical models, other fields (e.g., economics, psychology, sociology) often rely on statistical models due to numerous unknowns.
Economic Decisions: Statistical feedback is vital for informed decisions like interest rate changes, particularly during crisis scenarios (e.g., government shutdowns).

Statistical Outputs and Decisions: Statistical data provides a framework for decision-making but does not guarantee correctness.
Risk Assessment Examples: Calculation of audit probabilities based on income brackets and historical audit frequencies (calculation: average audits in bracket / total population).

Phase 1: Question Formulation: Determine the effect of different variables (e.g., homework frequency on student performance).
Phase 2: Experimental Design: Define the experimental setup (e.g., two exams versus four exams).
Phase 3: Data Collection: Establish methods to measure outcomes (e.g., average grades, retention rates).
Phase 4: Drawing Conclusions: Assess if changes yield expected educational outcomes (e.g., effectiveness of added exams).

Subjects/Units: Refers to participants in the study (e.g., class students).
Target Population: The broader group from which the subject group is drawn (e.g., all students in statistics classes).
Sample Frame: The specific class or section chosen for the study.
Sampling Method: Describes how the subjects are selected; emphasize lack of randomness due to choice in class preference.

Key Success Metrics: Assessment methods to determine success (e.g., percentage of students achieving each grade).
Single Factor Focus: Importance of isolating one variable to assess its specific impact, reducing potential confounding influences.
Data Comparison: Comparing success metrics across different sections or against historical data.

Random Selection: A method to reduce bias by ensuring all sections have equal chances of selection for the study.
Feedback and Citing Sources: Acknowledgment of intellectual contributions in research to maintain academic integrity.

Data Representation: Variables represented in columns for analysis; categorization of variables into categorical (qualitative) and quantitative types.
- Categorical Variables: Non-numeric categories (e.g., types of colors).
- Quantitative Variables: Numeric values representing measurements (e.g., test scores).
- Ordinal vs. Nominal: Distinction between values that can be sorted (ordinal) and those that cannot (nominal).

Standard Deviation: A measure that indicates the dispersion of data points from the mean.
- Interpretation: High standard deviation = values are spread out; low standard deviation = values clustered.

Data Analysis: Only basic analysis possible; frequency counting, bar charts, and pie charts are common methodologies.
Relativity of Averages: Interpretation and significance of average scores based on ordinal data (e.g., survey results).

Case Examples: Determining student success rates and survey results to provide actionable insights based on collected data.
Evolution of Categories through Research: Addressing the changes in variables over time and their influence on statistical outputs.

Experimental statistics provides a structured approach to testing hypotheses across myriad disciplines, particularly psychology and economics.
Importance of clear design, bias reduction, and proper parameter measurement in experiments to yield meaningful statistical insights.
Recognition of the complexities and variabilities inherent in human nature and behavior emphasizes the need for robust statistical applications in research.

Students are encouraged to engage with statistical analysis actively, bringing clarity to research questions while critically evaluating the impact of different studies on understanding human behavior.