PSYC250 Lecture 1 - Introduction to Statistics and Exploratory Data Analysis

This flashcard set covers the introductory materials of PSYC250, including campus safety protocols, the diagnostic method for learning statistics, assessment structures, and the fundamentals of Exploratory Data Analysis (EDA) including normality testing, boxplots, and confidence intervals.

SafeZone App

A free application for iOS, Android, and Windows providing Wellbeing Assistance, First Aid, or Emergency buttons for campus help.

SafetyNet

The University of Wollongong's online tool for reporting hazards and incidents.

UOW Smoke-Free Policy

A policy in effect since July 2016 that prohibits smoking, including the use of vapes and e-cigarettes, in all public areas, buildings, eating areas, grounds, pathways, and transport stops.

Ground Truth Notes

Personal, heavily edited notes provided by the lecturer on Moodle, specifically translated for inexperienced math learners.

Gap Protocol

The practice of documenting the exact moment a question occurs or a student loses the thread of understanding, referred to as finding your "X."

Pile A (Practice Test)

Questions answered correctly, quickly, and easily during a timed practice assessment.

Pile B (Practice Test)

Questions answered correctly but which took longer than the target pace (e.g., a mark a minute).

Pile C (Practice Test)

Questions that were attempted but answered incorrectly.

Pile D (Practice Test)

Questions where the student did not know how to begin the problem-solving process.

Pile E (Practice Test)

Questions that were left unaddressed because the student ran out of time during the assessment.

Exploratory Data Analysis (EDA)

The essential diagnostic phase used to check if a dataset meets the necessary assumptions required for a hypothesis test.

Null Hypothesis ($H_0$)

The hypothesis stating that a treatment has no effect and the sample mean will not differ significantly from the population mean (e.g., $\mu = 28$).

Alternative Hypothesis ($H_1$)

The hypothesis stating that a treatment has a significant effect and the sample mean will be significantly different from the population mean (e.g., $\mu \neq 28$).

Population

The large group of individuals that a scientific study is intended to apply to.

Sample

The specific, smaller group of people who participate in a study to generate scientific data.

Parameter

A numerical value that describes a population, typically represented by Greek letters such as $\mu$, $\sigma^2$, or $\sigma$.

Statistic

A numerical value that describes a sample, typically represented by Latin letters such as $M$ or $\bar{x}$, $s^2$, or $s$.

Standard Error

The standard deviation of the Sampling Distribution of the Mean, indicating how much the sample mean is expected to miss the true population mean.

Point Estimate

A single value, such as a sample mean, used as a best guess for the true population average.

95% Confidence Level

The accepted compromise in psychology between confidence and utility, meaning 95 out of 100 intervals created using this method are expected to capture the true population mean.

Z-score (multiplier)

A standardized measure of distance; in a normal distribution, taking approximately $1.96$ steps in both directions from the center captures 95% of the area.

Parametric Assumptions

The criteria required for parametric tests, including normally distributed data, interval data, and independent design.

Interval Data

Data where the distance between numbers represents equal amounts of the construct being measured, such as standardized test scores or reaction times.

Skewness

A measure of symmetry in a distribution; positive values indicate a pile-up on the left, while negative values indicate a pile-up on the right.

Kurtosis

A measure of the peakedness of a distribution; positive values indicate a pointy, heavy-tailed distribution, while negative values indicate a flat, light-tailed distribution.

Q-Q Plot

A Quantile-Quantile plot that maps observed data against a theoretical perfectly normal diagonal line to visually check for alignment with normality.

Interquartile Range (IQR)

The central 50% of the data, calculated as $Q3 - Q1$, representing the distance between the 25th and 75th percentiles.

Mild Outlier

An observation that falls between $1.5 \times IQR$ and $3 \times IQR$ away from the upper or lower hinges of a boxplot.

Extreme Outlier

An observation that falls more than $3 \times IQR$ away from the upper ($Q3$) or lower ($Q1$) hinges.

Shapiro-Wilk Test

A formal statistical test for normality recommended for small sample sizes where $N < 50$.

Kolmogorov-Smirnov Test

A formal statistical test for normality recommended for larger sample sizes where $N \ge 50$.

Significant Normality Test Result (p < .05)

Indicates that the sample distribution is significantly different from a normal distribution (i.e., non-normal).