Introduction to Statistics: Definitions, Data, and Design

For these estimates to be useful, the sample must be a representative sample, meaning it must contain the essential characteristics of the population. If a sample is biased—meaning some members of the population are more likely to be chosen than others—the resulting statistic will not accurately reflect the true population parameter.

Variables and Data Types

To conduct a study, we must identify a variable, which is a characteristic of interest for each member of a population. Variables are typically denoted by capital letters like $X$ or $Y$.

The actual values collected for these variables are called data. A single value is referred to as a datum. Data can be classified into two main types:

1. Numerical Variables: These take on values with equal units, such as weight, time, or GPA. Numerical data can be further divided into:
   • Discrete: Data resulting from counting (e.g., the number of books in a backpack).
   • Continuous: Data resulting from measuring (e.g., the weight of a backpack).
2. Categorical Variables: These place an item into a specific category, such as hair color or brand of car.

Levels of Measurement

Not all data are created equal. The "level of measurement" determines what kind of mathematical operations you can perform on the data:

• Nominal: Qualitative data used for labels or names (e.g., eye color). No mathematical calculation is possible.
• Ordinal: Data that can be ordered, but the differences between values cannot be measured (e.g., "excellent," "good," "satisfactory").
• Interval: Data with a definite ordering where differences can be measured, but there is no absolute zero (e.g., temperature in Celsius).
• Ratio: The highest level. It has a true zero point, allowing for the calculation of ratios (e.g., weight or exam scores).

Summarizing Data: The Arithmetic Mean

Once data is collected, we often look for a "center" to describe the group. The most common measure is the arithmetic mean, which is often referred to interchangeably as the average.

While "average" is common in everyday speech, "arithmetic mean" is the preferred technical term in statistics.

Experimental Design and Ethics

When researchers want to investigate if one variable causes a change in another, they design an experiment. In this setup, we have an explanatory variable (the cause) and a response variable (the effect).

To ensure accuracy, researchers use:

• Random Assignment: Distributing subjects to groups randomly to cancel out "lurking variables" (unintended factors that might influence results).
• Control Groups and Placebos: A group that receives no active treatment to account for the "power of suggestion".
• Blinding: Ensuring subjects (and sometimes researchers) do not know who is receiving the treatment to prevent bias.

Finally, ethics are paramount. Statistical fraud can occur when researchers manipulate data or stop collecting data prematurely to support a desired conclusion. True science relies on transparency and representative sampling to ensure that the "numbers" actually tell the truth.