Introduction to Statistics: Variables, Parameters, Descriptive and Inferential Statistics, and Sampling Error

Definition of a Variable: A variable is any characteristic, condition, or attribute that can change or assume different values across individuals or environments.
Classification of Variables:
- Individual Characteristics: Characteristics that naturally differ from one person to another. Examples include:
  - Weight
  - Gender identity
  - Personality traits
  - Motivation levels
  - Behavioral patterns
- Environmental Conditions: External contextual factors that vary across experimental settings or time. Examples include:
  - Ambient temperature
  - Time of day
  - Physical dimensions or size of a testing room
Environmental Impact Example (Weather and Restaurant Reviews):
- A study investigated whether external weather conditions influence customer online restaurant reviews.
- Variables: Environmental weather conditions and customer restaurant review ratings.
- Findings: A direct empirical relationship was demonstrated. Customer reviews were consistently worse during bad weather, specifically on days with extremely hot or extremely cold temperatures.
Measurement Terminology:
- Datum (Score / Raw Score): The single measurement or numerical value obtained for a specific individual in a study.
- Dataset (Data): The complete set or collection of measurements/scores gathered from all individuals participating in a research study.

Populations and Samples:
- Population: The entire, comprehensive group of individuals, objects, or items of interest in a research investigation (e.g., all registered voters).
- Sample: A specific subset selected from a population, intended to represent that population in a study (e.g., a group of middle school children).
- Populations and Samples of Scores: Because research involves measuring individuals to derive numerical values, every sample of individuals produces a corresponding sample of scores, and every population of individuals produces a corresponding population of scores.
Quantitative Descriptors:
- Parameter: A numerical characteristic or summary value that describes an entire population (e.g., the average score of a population).
- Statistic: A numerical characteristic or summary value that describes a sample (e.g., the average score of a sample).
Research Workflow:
- Investigations systematically begin with a research question concerning an unknown population parameter.
- Because populations are typically too large to measure every individual, researchers draw a representative sample.
- Sample data are measured to calculate sample statistics.
- Sample statistics serve as the practical foundation for answering questions and drawing inferences about population parameters.

Descriptive Statistics:
- Definition: Statistical procedures utilized to organize, simplify, and summarize raw data into a manageable, comprehensible form.
- Common Techniques:
  - Organizing raw scores into structured visual distributions using tables or graphs.
  - Calculating single summary values, such as an average, to represent an entire dataset containing hundreds of individual scores.
Inferential Statistics:
- Definition: Statistical methods that evaluate sample data to draw generalized conclusions, infer properties, and make statements about broader populations.

Concept of Sampling Error:
- Definition: The naturally occurring, unsystematic discrepancy or difference between a sample statistic and its corresponding population parameter.
- Mechanism: Samples provide limited information about a population. Because different samples consist of different individuals with unique attributes, sample statistics naturally vary from sample to sample and differ from the true population parameter.
- Inferential Challenge: Sampling error represents the fundamental problem that all inferential statistical methods must address.
Demonstration Example (Student Age Study):
- Population Parameters: A defined population of $1000$ college students with a true population mean age of $21.3\,\text{years}$ .
- Sample 1: A random sample of $5$ students selected from the population yields a mean age statistic of $19.8\,\text{years}$ .
- Sample 2: A second random sample of $5$ students selected from the same population yields a mean age statistic of $20.4\,\text{years}$ .
- Analysis: Both sample means differ from one another ( $19.8\,\text{years}$ vs. $20.4\,\text{years}$ ) and diverge from the population parameter ( $21.3\,\text{years}$ ). This discrepancy illustrates sampling error occurring across hundreds of possible random samples.

Political Polling and Margin of Error:
- Context: A political poll conducted on a sample of registered voters reports that Candidate Brown leads with $51\%$ of the vote, Candidate Jones holds $42\%$ approval, and $7\%$ of respondents are undecided.
- Margin of Error: The reported polling results include a margin of error of $\pm 4\%$ (plus or minus four percentage points).
- Interpretation: The reported sample percentages are generalized to the target population of all potential voters. The margin of error is the explicit quantification of sampling error.
Classroom Division Example:
- Procedure: A statistics classroom is partitioned into two groups by drawing an imaginary line from front to back through the middle of the room.
- Measurement: Computing the average age, height, or Grade Point Average ( $\text{GPA}$ ) for each side of the room.
- Outcome: The two groups will almost certainly yield different averages.
- Interpretation: An observed difference (such as a higher average age on the right side of the room) does not reflect a systematic force causing older students to sit on the right. Instead, it is the result of random, chance factors.
Core Role of Inferential Statistics: Inferential statistics determine whether observed differences between samples are caused by random chance factors (sampling error) or reflect genuine, meaningful relationships existing within the overall population.