Statistics and Probability Concepts

Foundation of Statistical Analysis: Built on probability theory, which deals with random outcomes.
- Example: Rolling a die is a classic case of randomness.
- Not All Events are Random: The weather is somewhat predictable but has random elements.
Probability: Likelihood of an event occurring. Even in predictable situations (like college admissions), random events can disrupt expectations.

Impact of Randomness in Life: Events can significantly affect decisions, although not always grounded in statistical likelihood.
- Example: A friend's fear post a carjacking led to irrational behavior regarding personal safety.
- Rational vs. Irrational Response: The friend’s response is a common cognitive bias; a single event can skew perception of risk.
Nature of Statistics: It allows measurement of likelihoods, enabling discussions about probabilities without claiming 100% certainty.

Point of the Story: Individual incidents don't reflect changes in the overall probability of events occurring.
- Example: A single crime incident doesn't mean increased likelihood of crime in a neighborhood.
Cognitive Bias: Humans tend to focus more on unusual events (availability heuristic) leading to irrational judgments.

Statistics and Population: Use statistics to infer characteristics about a larger population from a sample.
- Examples of Populations: U.S. citizens, GSU students, company sales data.
Limitations of Data: Difficulties in covering entire populations lead to reliance on samples for statistical inferences.
Confidence in Inferences: Probability theory helps gauge how confident we can be in our inferences based on sample data.

Importance of Representative Sampling: To ensure validity of inferences drawn about a population, samples must reflect population characteristics.
- Random Sampling: Ideal method for obtaining a representative sample.
- Limitations of Using Non-Representative Samples: Characteristics like age, geographic location, and socioeconomic status can skew results.
Consideration in Research Projects: Ensure the sample aligns with the population of interest to avoid flawed conclusions.

Generalizing Findings: Findings from a random sample can be generalized to the entire population if free from biases.
Concerns with Sample Limitations: Geographical biases, age group limitations, self-selection issues in survey responses can affect representativeness.
Self-Selection Problems: These affect generalizability of findings, leading to potential misconceptions as demonstrated in dissertation research examples.

Thinking About Research Questions: Statistical tests are used to explore expected differences or relationships between variables.
Hypothesis Testing: Importance of formulating expectations as hypotheses to evaluate through statistical probability.
Understanding p-values: Reflects risk of making incorrect conclusions from statistical analysis; no conclusion can be 100% certain due to inherent margin for error.
Aim for High Confidence: While 99% confidence is substantial, a completely certain conclusion is unattainable in statistical analysis.