Descriptive Statistics and Data Collection Methods in Health Sciences

Introduction to Statistics

Why Statistics Matter in Health Sciences

Statistics is essential for transforming health data into meaningful evidence. It enables researchers to test treatments, identify risk factors, and evaluate research outcomes with confidence. By mastering statistical methods, students gain critical thinking skills and the ability to make evidence-based decisions that improve health outcomes.

Populations and Samples

Definitions and Relationships

A population is the entire group of individuals of interest in a study, while a sample is a subset of the population selected for analysis. The goal is to use sample data to make inferences about the population.

• Population: All individuals of interest (e.g., all students at a university).

• Sample: Individuals selected to participate in the research study.

• Inferential Statistics: Using sample data to draw conclusions about the population.

Example: Population vs. Sample

In a survey of 250 students at the University of Ottawa, 35 reported smoking regularly. The population is all students at the university, while the sample is the 250 surveyed students.

Parameters and Statistics

Key Definitions

• Parameter: A numerical description of a population characteristic (e.g., average age of all students).

• Statistic: A numerical description of a sample characteristic (e.g., average age of surveyed students).

Example

• If a survey of 450 students reports an average weekly income of $325, this is a sample statistic.

• If the average weekly income for all students is $405, this is a population parameter.

Sampling Error

Definition and Importance

Sampling error is the discrepancy between a sample statistic and its corresponding population parameter. It arises naturally due to chance and is a central concept in inferential statistics.

• Sampling error quantifies the uncertainty in generalizing sample results to the population.

• Measuring and minimizing sampling error is crucial for reliable statistical inference.

Methods of Sample Collection

Simple Random Sampling (SRS)

In SRS, every possible sample of size n from a population of size N has an equally likely chance of being selected. This method ensures unbiased representation.

• Random number generators or tables are often used to select samples.

Systematic Sampling

Systematic sampling involves selecting a starting point and then choosing every kth element in the population.

• Example: Selecting every 3rd person from a list.

Stratified Sampling

Stratified sampling divides the population into subgroups (strata) that share specific characteristics, then samples are drawn from each subgroup.

• Ensures representation of all subgroups (e.g., gender, age).

Cluster Sampling

Cluster sampling divides the population into clusters, randomly selects some clusters, and includes all members from selected clusters.

• Useful for geographically dispersed populations.

Variables in Research

Definition and Types

A variable is any characteristic or condition that can change or take on different values. Research often investigates relationships between two or more variables.

• Discrete Variables: Take on a finite or countable number of values (e.g., number of patients).

• Continuous Variables: Can take on infinitely many values within a range (e.g., weight, blood pressure).

Levels of Measurement

Overview

The level of measurement determines which statistical calculations are meaningful. There are four levels: nominal, ordinal, interval, and ratio.

Level	Description	Examples
Nominal	Qualitative; categories only, no order, no math	Colors, names, types
Ordinal	Ordered categories, but differences not meaningful	Class standings, rankings
Interval	Ordered, equal intervals, no true zero	Temperature, years
Ratio	Ordered, equal intervals, true zero	Height, weight, length

Nominal Level

Nominal data are qualitative and consist of names, labels, or categories. No mathematical computations are possible.

• Examples: Colors in a flag, student names, textbook titles.

Ordinal Level

Ordinal data are qualitative and can be arranged in order, but differences between entries are not meaningful.

• Examples: Class standings, player numbers, song rankings.

Interval Level

Interval data are quantitative, ordered, and have equal intervals, but zero does not represent the absolute lowest value.

• Examples: Temperatures (Fahrenheit), years on a timeline.

Ratio Level

Ratio data are quantitative, ordered, have equal intervals, and a true zero. Zero means none of the variable exists.

• Examples: Height, weight, length.

Key Distinctions

• Nominal: Categories only, no order, no math

• Ordinal: Ordered categories, no meaningful differences

• Interval: Ordered, equal intervals, no true zero

• Ratio: Ordered, equal intervals, true zero

Summary Table

Concept	Definition
Population	Entire group of interest
Sample	Subset of population
Parameter	Numerical value describing population
Statistic	Numerical value describing sample
Sampling Error	Difference between statistic and parameter
Simple Random Sample	Every sample equally likely
Systematic Sampling	Select every kth element
Stratified Sampling	Sample from each subgroup
Cluster Sampling	Sample all members from selected clusters
Discrete Variable	Finite/countable values
Continuous Variable	Infinite values within range
Nominal	Categories only
Ordinal	Ordered categories
Interval	Ordered, equal intervals, no true zero
Ratio	Ordered, equal intervals, true zero

Key Formulas

Sampling Error Formula

The sampling error is calculated as:

$\text{Sampling Error} = \text{Sample Statistic} - \text{Population Parameter}$

Probability of Simple Random Sample

For a population of size N and sample size n:

$\text{Probability of a specific sample} = \frac{1}{\binom{N}{n}}$

Conclusion

Understanding populations, samples, sampling methods, variables, and levels of measurement is foundational for descriptive statistics and for conducting meaningful health science research. These concepts enable researchers to collect, analyze, and interpret data accurately, ensuring valid and reliable conclusions.