Inferential Statistics: Population, Sample, and Parameters

---

Population and Sample: Definitions and Distinction

---

In statistics, we are often interested in understanding characteristics of a large group of individuals or items. It is crucial to distinguish between the entire group of interest and the smaller subset from which we typically collect data.

Population

• Definition:

  A **population** refers to the **entire collection** of individuals, objects, events, or measurements that share a common characteristic and are of interest to a statistical study. It is the complete set about which we want to draw conclusions or make inferences.

• Scope:

  It includes every single member of the group being studied. For example, if studying the heights of students in a specific school, the population is ALL students currently enrolled in that school.

• Nature:

  Populations can be **finite** (countable, like all residents of a city) or **infinite** (uncountable, like all possible outcomes of rolling a die indefinitely, or all potential pressure readings in a continuous process).

• Goal:

  The ultimate goal of most statistical studies is to describe or make inferences about the characteristics of the population.

---

Sample

• Definition:

  A **sample** is a **subset or subgroup** selected from the population. It is a smaller, manageable collection of individuals or items from which data is collected and analyzed.

• Purpose:

  Collecting data from an entire population (conducting a census) is often impractical due to limitations in time, cost, resources, or accessibility. Sampling allows us to gather data from a smaller, representative subset of the population to draw conclusions about the larger group.

• Representativeness:

  For the inferences drawn from the sample to be valid and generalizable to the population, the sample must be **representative** of the population. This means the sample should reflect the characteristics of the population as accurately as possible. Using appropriate sampling techniques (discussed implicitly in the next section) is crucial for achieving representativeness and minimizing bias.

---

Distinction and Relationship

The key difference between a population and a sample lies in their scope:

• The **population** is the **entire** group of interest.
• The **sample** is a **part** of that group from which data is collected.

The process of **statistical inference** involves using information obtained from the sample (statistics) to make generalizations or draw conclusions about the characteristics of the population (parameters). We collect data from the sample because we cannot usually access the entire population.

---

Examples

Scenario	Population	Sample
Investigating the average lifespan of a specific brand of LED bulbs	All LED bulbs of that specific brand that have been or will be produced. (Potentially infinite)	A set of 500 LED bulbs of that brand tested under controlled conditions.
Determining the opinion of voters on a new policy	All individuals who are registered and eligible to vote in the relevant election/area. (Finite, but often very large)	A group of 1500 registered voters contacted and surveyed by a polling agency.
Studying the effectiveness of a teaching method in secondary schools in a state	All secondary school students in that specific state. (Finite)	Students in 5 schools in that state who participate in a study using the new method.
Analyzing the blood sugar levels of patients with diabetes	All individuals diagnosed with diabetes. (Potentially very large, can be considered infinite for practical purposes)	Blood sugar measurements from 250 diabetic patients participating in a study.

---

Parameter and Statistics: Definitions

---

Numerical summaries are used to describe characteristics of both populations and samples. However, different terms are used depending on whether the measure describes the entire population or just the sample.

Parameter

• Definition:

  A **parameter** is a numerical value that describes a characteristic of the entire **population**. It is a fixed value that is typically unknown because we usually do not have data for every single member of the population.

• Nature:

  Parameters are constant values for a given population, although their exact values are usually what we are trying to estimate or make inferences about in a statistical study.

• Notation:

  Parameters are conventionally denoted by Greek letters.

Common Population Parameters:

• **Population Mean:** The average value of a variable for all individuals in the population, denoted by $\mu$ (mu).
• **Population Standard Deviation:** A measure of the spread or variability of the variable's values across the entire population, denoted by $\sigma$ (sigma).
• **Population Variance:** The square of the population standard deviation, $\sigma^2$.
• **Population Proportion:** The proportion (or percentage) of individuals in the population who possess a specific characteristic, often denoted by $p$ or $\pi$.
• **Population Correlation Coefficient:** Measures the linear relationship between two variables in the population, denoted by $\rho$ (rho).

---

Statistic

• Definition:

  A **statistic** is a numerical value that describes a characteristic of a **sample**. It is calculated directly from the data collected from a sample.

• Nature:

  Statistics are known values once the sample data is collected. However, the value of a statistic can vary from sample to sample, even if the samples are drawn from the same population using the same method. This variability is known as sampling variability.

• Purpose:

  Statistics are used as estimates of unknown population parameters. For example, the sample mean is used to estimate the population mean.

• Notation:

  Statistics are conventionally denoted by Roman letters.

Common Sample Statistics:

• **Sample Mean:** The average value of a variable calculated from the sample data, denoted by $\bar{x}$ ("x-bar"). Used to estimate $\mu$.
• **Sample Standard Deviation:** A measure of the spread of the variable's values within the sample, denoted by $s$. Used to estimate $\sigma$.
• **Sample Variance:** The square of the sample standard deviation, $s^2$. Used to estimate $\sigma^2$. (Note: Sample variance for inference often uses a denominator of $n-1$ instead of $n$ in its calculation).
• **Sample Proportion:** The proportion (or percentage) of individuals in the sample who possess a specific characteristic, often denoted by $\hat{p}$ ("p-hat") or $\bar{p}$. Used to estimate population proportion $p$.
• **Sample Correlation Coefficient:** Measures the linear relationship between two variables in the sample, denoted by $r$. Used to estimate $\rho$.

---

Relationship

The core idea of inferential statistics is to use statistics calculated from a sample to make inferences about parameters of the population from which the sample was drawn.

Because a sample is typically only a small part of the population, a statistic is an estimate of the parameter and will not be exactly equal to the parameter, although a good statistic from a representative sample should be close.

---

Example

Answer:

---

Sampling Techniques (Types of Sampling - Implicit)

---

Purpose of Sampling

As established, directly studying an entire population is often impractical, expensive, or impossible. Therefore, researchers rely on collecting data from a sample to gain insights into the characteristics of the larger population. The fundamental objective of sampling is to select a sample that is **representative** of the population. A representative sample accurately reflects the relevant characteristics of the population from which it was drawn.

If a sample is not representative, it is considered **biased**. Conclusions drawn from a biased sample may not accurately reflect the population and can lead to incorrect or misleading results.

---

Importance of Sampling Method

The **method** used to select the sample is critically important in determining whether the sample is representative and, consequently, in determining the validity and reliability of statistical inferences made about the population. A well-designed sampling method aims to minimize sampling bias and provide a basis for estimating sampling error (the natural variability of statistics from sample to sample).

Sampling methods are broadly classified into two main categories:

• Probability Sampling Methods:

  In these methods, every member of the population has a known, non-zero probability of being selected for the sample. The selection process involves some form of random mechanism, which helps to ensure representativeness and allows researchers to use probability theory to make inferences about the population and quantify the reliability of these inferences.

  Examples of Probability Sampling Techniques:

  • Simple Random Sampling (SRS): Every possible sample of a given size has an equal chance of being selected. Like drawing names from a hat or using a random number generator.
  • Systematic Sampling: Selecting individuals at regular intervals from a sampling frame (e.g., selecting every 10th person from a list).
  • Stratified Random Sampling: Dividing the population into mutually exclusive subgroups (strata) based on relevant characteristics (e.g., age groups, gender) and then drawing a simple random sample from each stratum.
  • Cluster Sampling: Dividing the population into clusters (e.g., geographic areas, schools) and then randomly selecting some clusters and including all members within the selected clusters in the sample.

• Non-Probability Sampling Methods:

  In these methods, the selection of the sample is not based on a random process. The probability of selecting any particular member is unknown. These methods are often used for convenience, speed, or cost savings, but they are more prone to sampling bias, and formal statistical inference to the population is generally not valid or should be interpreted with extreme caution.

  Examples of Non-Probability Sampling Techniques:

  • Convenience Sampling: Selecting individuals who are easily accessible (e.g., surveying people on a street corner).
  • Judgmental (Purposive) Sampling: The researcher selects individuals based on their own judgment about who would be representative or knowledgeable.
  • Quota Sampling: Sampling until a specific number of individuals from different subgroups are included, but without using a random selection process within subgroups.
  • Snowball Sampling: Participants recruit other participants from among their acquaintances. Used for hard-to-reach populations.

While the detailed mechanics of each technique are extensive, the overarching principle in inferential statistics is that **probability sampling methods**, especially those incorporating randomness like Simple Random Sampling, are preferred because they are designed to produce representative samples and provide the necessary theoretical basis for making statistically valid inferences from the sample to the population.