Class Notes on Experimental Design and Sampling Techniques
Class Structure and Homework Assignments
- Students are given class time to complete homework.
- Homework is due in one week if not finished in class.
- Chapter one homework is likely due on Friday of next week.
- Students are encouraged to stay up to date with homework to prevent last-minute stress.
Attendance and Student Engagement
- Instructor checks attendance by calling out names of students present in class.
- Students are reminded about a bonus point system for attendance.
- Emphasis on participation and being engaged.
Winter Weather Considerations
- Mention of an approaching snowstorm (referred to as "snowmageddon").
- No Zoom lectures will be conducted as they are deemed awkward and ineffective.
- Discussion about adapting to weather changes without relying on digital lectures.
Overview of Experimentation
Main Concepts in Experiments
- What Makes a Good Experiment: Randomization is a key factor.
- Definition of Randomness: Everything or every subject must have an equal chance of selection for treatment groups.
Example of Randomization
- Example discussed about testing two medicines: one real and one placebo.
- Random assignment is needed to ensure that results are valid.
Potential Issues with Randomization
- Risk of unbalanced groups, especially regarding characteristics such as age.
- Older individuals may require longer recovery from illness, leading to potential bias if all are placed in one treatment group.
Solution: Randomized Block Design
- Utilize a technique called randomized block design to ensure balanced representation across treatment groups.
- Each group must contain a representative sample, preventing age-related bias.
- Example scenario: Ensuring equal numbers of older subjects in both treatment groups.
Importance of Replication
- Replication is crucial for validating experimental results.
- Results must be replicable by others using the same methods.
- Example provided concerning the Innocence Project involving DNA testing: all DNA samples must leave enough material to allow for retesting.
Statistical Concepts
- Transition from discussing populations to samples within statistics.
- Introduction of common sampling methods that help researchers obtain valid data.
- Four primary sampling methods will be detailed below.
Common Sampling Methods
1. Simple Random Sample (SRS)
- Definition: A simple random sample is drawn from the entire population with every individual having an equal chance of selection.
- Example method for random selection could involve drawing names from a hat.
Electronic Random Selection
- Example use of Excel function
=RANDBETWEEN(lower limit, upper limit) to randomly select individuals or items from a defined population. - Importance of using unique identifiers (e.g., student ID numbers) for accurate random selection.
Advantages of SRS
- Reduces bias significantly by ensuring diversity within the sample.
Bias Consideration in Simple Random Sampling
- Acknowledgment of the chance that a sample could still be biased despite randomness due to variable grouping (i.e., all individuals from one side of the classroom).
2. Stratified Random Sample
- Definition: In stratified sampling, the population is divided into subgroups (strata) and a random sample is taken from each stratum.
- Example of strata could include characteristics like major, living situation, or year of study.
Advantages of Stratification
- Reduces bias more effectively than simple random sampling as it ensures all groups are represented.
Limitations of Stratified Sampling
- More time-consuming and complex to implement compared to simple random sampling.
3. Cluster Sample
- Definition: Instead of randomly selecting individuals from the entire population, whole clusters (groups) are selected at random.
- Example: Selecting entire classrooms or dorms to represent a larger population.
Risks of Cluster Sampling
- Potentially introduces bias if selected groups do not represent the overall population.
- Example discussed involving a study focusing on one fraternity to assess campus party culture.
4. Systematic Sample
- Definition: Using a systematic method to select individuals, often by choosing every n-th member from a list.
- Example: If selecting every third person from a list of class members during a survey.
Application of Systematic Sampling
- Practical for situations like cafeteria reviews where multiple groups of people are surveyed systematically instead of randomly.
Convenience Sampling
- Definition: A method where the sample is taken from a group that is easy to access or sample.
- Potential Issues with Bias: Convenience sampling is likely to introduce bias as it does not represent the larger population well and often lacks randomization.
- Example provided regarding media coverage of public opinions at diners, which might not reflect broader public sentiments.
Critical Factors Impacting Statistical Studies
1. Law of Small Numbers
- Refers to drawing conclusions about a population based on a small sample size.
- Example: A study sampling one classroom out of thousands and generalizing the results to the entire student population.
2. Confounding Variables
- Issues arise from not accounting for external factors that might influence the results of a study, such as socioeconomic status influencing intelligence, which should be taken into account in research.
3. Data Manipulation
- Ethical concerns raised regarding intentional or unintentional data manipulation leading to false conclusions in research.
Important Ethical Considerations in Research
- Ensure that participants are informed of all relevant risks when partaking in studies.
- Analysis of historical research indicating failure to disclose risks and implications (e.g., various unethical experiments conducted in the past).
Wrap-Up and Future Tasks
- Students will work in small groups for chapter one investigations to analyze data and implement sampling techniques learned.
- Emphasis on collaboration in groups of two to five students for completing future assignments effectively.
Homework and Projects
- Encouragement to start homework assignments promptly to avoid last-minute issues.
- Clear expectations set for group work and future investigations in class on Friday.