Introduction to Statistics: Key Concepts and Methods

Introduction to Statistics

What is Statistics?

Statistics is the science of planning studies and experiments, obtaining data, and organizing, summarizing, presenting, analyzing, and interpreting those data to draw conclusions. It is essential for making informed decisions based on data.

• Data: Collections of observations, such as measurements, genders, or survey responses. A single data value is called a datum.

• Population: The complete collection of all measurements or data being considered.

• Census: The collection of data from every member of the population.

• Sample: A subcollection of members selected from a population.

• Parameter: A numerical measurement describing some characteristic of a population.

• Statistic: A numerical measurement describing some characteristic of a sample.

Statistical Thinking and Sampling

Sound statistical analysis requires representative sample data and careful attention to methodology. Poor sampling methods can lead to biased or useless results.

• Representative Sample: Sample data must accurately reflect the population.

• Random Selection: Essential for obtaining unbiased, representative samples.

• Voluntary Response Samples: (e.g., internet, mail-in, or call-in polls) are often biased and should be avoided for making population inferences.

• Statistical Significance: Achieved when a result is very unlikely to occur by chance (commonly, probability ≤ 5%).

• Correlation vs. Causation: Correlation does not imply causation; observed associations may not reflect direct cause-effect relationships.

Common Issues in Data Collection

• Loaded Questions: Poorly worded survey questions can bias results.

• Order of Questions: The sequence of questions can influence responses.

• Nonresponse: Occurs when individuals refuse or are unavailable to respond, potentially biasing results.

• Percentages: Always interpret percentages as "divided by 100." Be cautious of misleading or unclear percentage claims.

Calculating Percentages

• Percentage of a Value: Drop the percent sign, divide by 100, then multiply by the whole value. Example: 6% of 1200 = $6 \div 100 \times 1200 = 72$

• Decimal to Percentage: Multiply by 100%. Example: $0.25 \times 100\% = 25\%$

• Fraction to Percentage: Divide numerator by denominator, then multiply by 100%. Example: $\frac{3}{4} = 0.75 \rightarrow 0.75 \times 100\% = 75\%$

• Percentage to Decimal: Divide by 100. Example: $85\% = 85 \div 100 = 0.85$

Types of Data

Quantitative vs. Categorical Data

• Quantitative (Numerical) Data: Numbers representing counts or measurements (e.g., age in years).

• Categorical (Qualitative) Data: Names or labels (e.g., eye color, gender, identification numbers used as labels).

Discrete and Continuous Data

• Discrete Data: Quantitative data with finite or countable values (e.g., number of physical exams given in a week).

• Continuous Data: Quantitative data with infinitely many possible values, not countable (e.g., volume of blood drawn between 0 mL and 50 mL).

Levels of Measurement

• Nominal: Categories only, no order or numerical significance (e.g., eye color).

• Ordinal: Categories with a meaningful order, but differences between values are not meaningful (e.g., letter grades).

• Ratio: Ordered, meaningful differences and ratios, with a natural zero (e.g., heights, weights).

Examples of Data Types

• Quantitative Data: Ages of subjects in a clinical trial.

• Categorical Data as Labels: Genders of subjects in a clinical trial.

• Categorical Data as Numbers: Identification numbers assigned to subjects.

• Discrete Data (Finite): Number of physical exams given in a week.

• Discrete Data (Infinite): Number of die rolls until a 5 appears.

• Continuous Data: Volume of blood drawn from a patient.

Big Data and Data Science

Big Data

Big data refers to data sets so large and complex that traditional software tools cannot analyze them efficiently. Analysis may require parallel processing on many computers.

• Examples of Big Data Magnitudes:

  • Terabytes ($10^{12}$ bytes)

  • Petabytes ($10^{15}$ bytes)

  • Exabytes ($10^{18}$ bytes)

  • Zettabytes ($10^{21}$ bytes)

  • Yottabytes ($10^{24}$ bytes)

Data Science: Involves statistics, computer science, software engineering, and other fields to analyze and interpret complex data sets.

Handling Missing Data

Types of Missing Data

• Missing Completely at Random: The likelihood of a value being missing is independent of its value or other values.

• Missing Not at Random: The missing value is related to the reason it is missing (e.g., low-income individuals not reporting salary).

Methods for Handling Missing Data

• Delete Cases: Remove all subjects with missing values. Appropriate if data are missing completely at random; otherwise, may introduce bias.

• Impute Missing Values: Substitute missing values using methods such as the mean, random selection from similar cases, or regression analysis.

When analyzing data with missing values, determine why values are missing and whether the remaining data are representative of the population.

Sampling Methods

Types of Sampling

• Simple Random Sample: Every possible sample of the same size has the same chance of being chosen.

• Systematic Sample: Select every kth subject (e.g., every 50th person).

• Convenience Sample: Use data that are easy to obtain (often biased).

• Stratified Sample: Subdivide the population into strata (groups) with similar characteristics, then randomly sample within each stratum.

• Cluster Sample: Partition the population into clusters, randomly select some clusters, then select all members from those clusters.

Types of Studies

• Cross-Sectional Study: Data are collected at one point in time.

• Retrospective (Case-Control) Study: Data are collected from past records or interviews.

• Prospective (Cohort) Study: Data are collected in the future from groups sharing common factors.

Experimental Design

Key Concepts

• Experiment: Apply a treatment and observe its effects on subjects (experimental units).

• Observational Study: Observe and measure characteristics without modifying subjects.

• Replication: Repetition of an experiment on multiple subjects to observe treatment effects.

• Blinding: Subjects do not know whether they receive treatment or placebo, reducing placebo effect.

• Randomization: Assign subjects to groups by random selection to create similar groups.

• Confounding: When the effect of one variable cannot be distinguished from another.

Experimental Designs

• Completely Randomized Design: Assign subjects to treatment groups by random selection.

• Randomized Block Design: Form blocks (groups) of similar subjects, then randomly assign treatments within each block.

Calculator Use in Statistics

TI-83/84 Plus Calculators

This course references the TI-84 Plus CE, TI-84 Plus C Silver Edition, TI-84 Plus Silver Edition, TI-84 Plus, and TI-83 Plus calculators. Instructions and screen displays are based on the TI-84 Plus CE, but all listed calculators are similar in function unless otherwise noted.

• Practice using your calculator for statistical computations as described in your textbook or course materials.

Summary Table: Types of Data and Sampling Methods

Type	Description	Example
Quantitative Data	Numerical values representing counts or measurements	Age in years, height in cm
Categorical Data	Names or labels, not numerical	Eye color, gender
Discrete Data	Countable values, finite or infinite	Number of exams, die rolls
Continuous Data	Infinitely many values, not countable	Blood volume drawn, temperature
Simple Random Sample	Every sample of size n has equal chance	Randomly select 10 students from a class
Systematic Sample	Select every kth subject	Every 5th person on a list
Stratified Sample	Divide into strata, sample within strata	Sample males and females separately
Cluster Sample	Divide into clusters, sample all in selected clusters	Sample all students in selected classrooms