Chapter 1: Introduction to Statistics – Collecting Sample Data

Introduction to Statistics

Statistical and Critical Thinking

Statistics is the science of collecting, analyzing, interpreting, and presenting data. Critical thinking is essential in statistics to ensure that data collection and analysis are valid and reliable. The method used to collect sample data directly influences the quality of statistical analysis. If data are not collected appropriately, the results may be invalid, regardless of the analysis performed.

• Key Concept: The simple random sample is of particular importance in statistics because it ensures that every possible sample of a given size has an equal chance of being selected.

• Importance: Poor data collection methods can render data useless, even with advanced statistical techniques.

Types of Data Collection

Data for statistical analysis are typically obtained from two main sources: observational studies and experiments.

Observational Study

An observational study involves observing and measuring specific characteristics without attempting to modify the individuals being studied.

• Purpose: To identify associations between variables.

• Limitation: Cannot establish causation, only association.

• Example: A study comparing the ages at death of left-handed and right-handed people by observing existing data without intervention.

Experiment

An experiment involves researchers imposing treatments and controls, then observing characteristics and taking measurements.

• Experimental Units: The individuals on whom the experiment is performed. When these are people, they are called subjects.

• Purpose: To establish causation by manipulating variables and observing effects.

• Example: A randomized trial assigning women to receive aspirin or placebo to study cardiovascular outcomes.

Design of Experiments

Proper experimental design is crucial for obtaining valid and reliable results. Key elements include replication, blinding, and randomization.

Replication

• Definition: Repetition of an experiment on more than one individual.

• Purpose: To ensure that results are not due to chance and to observe the effects of treatments across a larger sample.

Blinding

• Definition: A technique in which the subject does not know whether they are receiving a treatment or a placebo.

• Purpose: To prevent the placebo effect, where subjects report improvement due to expectations rather than the treatment itself.

Double-Blind

• Definition: Both the subject and the experimenter do not know whether the subject is receiving the treatment or placebo.

• Purpose: To eliminate bias from both participants and researchers.

Randomization

• Definition: Assigning subjects to different groups through a process of random selection.

• Purpose: To create groups that are similar and reduce selection bias.

Sampling Methods

Sampling is the process of selecting a subset of individuals from a population to estimate characteristics of the whole population. Several sampling methods are used in statistics:

Simple Random Sample

• Definition: A sample of n subjects is selected so that every possible sample of the same size has the same chance of being chosen.

• Note: Sometimes called a "random sample," but strictly, a random sample only requires that all members have the same chance of being selected.

• Example: Using a random number generator to select students from a class roster.

Systematic Sampling

• Definition: Select a starting point and then select every kth element in the population.

• Example: Selecting every third student from a list.

Convenience Sampling

• Definition: Use data that are very easy to obtain.

• Limitation: May introduce significant bias and is generally not recommended for rigorous statistical analysis.

Stratified Sampling

• Definition: Divide the population into at least two different groups (strata) that share similar characteristics, then take a sample from each group.

• Example: Dividing a class into males and females, then randomly selecting students from each group.

Stratified Random Sampling with Proportional Allocation

• Steps:

  1. Divide the population into subpopulations (strata).

  2. From each stratum, obtain a simple random sample of size proportional to the stratum's size.

  3. Use all selected members as the sample.

• Formula:

$\text{Sample size for stratum} = \frac{\text{Total sample size} \times \text{Stratum size}}{\text{Population size}}$

• Example: If a population of 1000 is divided into strata of sizes 300, 200, 400, and 100, and a sample of 20 is needed, the sample from each stratum is calculated proportionally.

Cluster Sampling

• Definition: Divide the population area into sections (clusters), randomly select some clusters, and choose all members from those selected clusters.

• Example: Randomly selecting a row in a classroom and surveying all students in that row.

Multistage Sampling

• Definition: Collect data using a combination of sampling methods in different stages.

• Example: Pollsters may use stratified sampling at the first stage and cluster sampling at the second stage.

Comparison of Sampling Methods

Sampling Method	Definition	Example
Simple Random Sample	Every possible sample of the same size has an equal chance of being chosen	Randomly select 6 students from a class roster
Systematic Sample	Select every k-th member from a list	Select every 3rd student from a list
Stratified Sample	Divide population into strata and sample from each	Divide by gender, select 3 males and 3 females
Cluster Sample	Divide into clusters, randomly select clusters, sample all in selected clusters	Select a row in a classroom, survey all students in that row

Applications and Examples

• Observational Study Example: Studying the age at death of left-handed vs. right-handed people by observing existing records.

• Experiment Example: Randomly assigning women to receive aspirin or placebo to study the effect on cardiovascular events.

• Sampling Example: To obtain a simple random sample of 6 students from a class, assign numbers to all students and use a random number generator to select 6.

Additional info: In practice, the choice of sampling method depends on the research question, available resources, and the need to minimize bias and maximize representativeness.