Statistics in Psychology

Statistics as a Tool:
- Statistics provide methods to organize and analyze collected data.
- Different types of statistics present data in an understandable form.
- Knowledge of statistics aids in interpreting arguments based on data.
- Statistics can reveal insights into societal or organizational problems.
- Example: Evaluating customer satisfaction with products.
Applications of Statistics:
- Summarizing data (e.g., average customer satisfaction).
- Examining score variability (e.g., range of scores).
- Displaying data via graphs and tables for better understanding.
- Testing predictions on average scores across different groups or times.
- Assessing relationships between datasets (e.g., customer satisfaction vs. information accessibility).

Statistics appear in various media sources and everyday contexts.
- Example: Colleges often share statistics regarding student success.
- Statistical Claim: "95% of Mizzou graduates have successful career outcomes within six months of graduation."
Media frequently uses statistics to:
- Present information about reported subjects.
- Advocate for specific viewpoints.

An analysis of teenage smoking rates can highlight patterns and inform regulations.

Observational data can lead to conclusions about the dangers of using cell phones while driving.

Statistics can be illustrated in multiple formats.
Awareness of potential biases in statistical presentation can help identify misleading arguments.
- Example: Claims linking excessive cellphone use to higher brain cancer rates need context.
Important considerations before making decisions based on stats:
- Magnitude of cancer rate increase.
- Methodology of the study conducted.
- Variances between groups in the study.
- Statistical significance of differences observed.
Definition of Statistically Significant: A statistical outcome indicating that data suggest an effect or relationship exists.

Statistics in research stem from studies aimed at understanding human behavior.
- Example: A study on how cell phone use impacts driving performance (Drews et al., 2009).

Definition: Type of statistics that summarizes or describes data collected.
- Elements:
- Mean: The average score within a dataset.
- Frequency: The occurrence rate of specific responses.
- Variability: The range/spread of scores in a distribution.
Example from Research:
- Study by Drews et al. found participants had delayed responses to brake lights and increased crash rates while texting.

Definition: Type of statistics that tests hypotheses regarding the data.
- Example Question: Does driving while texting lead to poorer driving performance compared to normal driving?

Research generates new knowledge across scientific fields by investigating varied phenomena.
- Example: Assessing how anxiety levels in college students relate to exam performance.

The nature of research questions determines the required research design.
- Variable Definition: Attributes that can change across individuals.
- Examples include height, age, temperature, and test scores.

Research observation methods impact how data statistics are interpreted.
The selection of methodology reflects the research question posed by the investigator.

**Experiments:
- Involve comparison of observed behaviors under varying situations outlined by an independent variable.**
- Independent Variable (IV): Factors manipulated (e.g., time spent studying).
- Dependent Variable (DV): Behavioral outcome studied (e.g., grade in the class).
- Goal: To examine behavior changes in response to differing situations.
- True Experiment: Participants randomly assigned to specific conditions (e.g., study groups).
- Quasi-Experiment: Participants are pre-existing in various conditions (e.g., comparing two sections of a course).
Between-Subjects Design:
- Different groups experience varying conditions.
- Example: Testing stress reduction programs among different groups of participants.
Within-Subjects Design:
- Same group experiences all conditions allowing individual comparisons over time.
- Example: Surveying mental health in undergraduates over weeks.

Definition: Evaluate relationships between different behavioral measures.
Goal: To identify if two behaviors are correlated.
Limitations: Correlation does not imply causation.
- Relationships could be direct (one causing the other) or indirect (third factor influencing both).
**Types of Correlations:
- Positive Correlation:** Both variables increase together.
- Negative Correlation: One variable increases while the other decreases.
Graphical representation helps envision correlation types.

Validity: The degree to which a study effectively tests the intended hypothesis.
- Types:
- Internal Validity: Shows causal relationships from the study.
- External Validity: How applicable findings are to broader contexts.
- Construct Validity: Accuracy of a measure to capture the behavior of interest.

Reliability: Consistency in behavioral measures.
Evaluated through statistical relationships:
- Inter-Rater Reliability: Agreement between different observers measuring similar behaviors.
- Test-Retest Reliability: Stability of scores when repeated under similar conditions.
- Internal Consistency: Testing consistency among different items within a survey.