Research Questions and Experimental Design

Four Types of Research Questions

1. Making an Estimate

   • Example: Estimating the average age of statistics students in California.
   • Method: Take a sample of students.

2. Testing a Claim

   • Example: Testing if the average age of statistics students in California is greater than 19 when research shows it is exactly 19.
   • Method: Hypothesis testing to verify the claim.

3. Comparing Two Populations

   • Example: Comparing statistics success rates between students at Napa Valley College and Solano College.
   • Method: Estimate or test the differences between the populations.

4. Investigating a Relationship

   • Example: Relationship between the number of hours watching study videos and course grades.
   • Method: Analyze the correlation between variables.

Components of Research Questions

1. Population

   • Example: Napa Valley College students.

2. Variable

   • Example: Attendance status (full-time or not).

3. Numerical Characteristic

   • Example: Proportion or percentage of students attending full-time.

Identifying Research Question Components

• Example Question: Do the majority of Napa Valley College students attend full-time?
  • Population: Napa Valley College students.
  • Variable: Attendance status (full-time or not).
  • Numerical Characteristic: Percentage of full-time students.

Observations vs. Experiments

Cause-and-Effect Relationships

• Explanatory Variable: Variable that is changed to measure its effects.
• Response Variable: Result of the change in the explanatory variable.

Types of Studies

• Observational Study

  • No variables are manipulated.
  • Example: Recording hours of video watched without altering study habits.
  • Keywords: Connected, associated, correlated, likelihood.

• Experimental Study

  • Variables are manipulated to observe effects.
  • Example: Testing the effect of different cooking temperatures on turkey doneness.

Explanatory vs. Response Variables

• Explanatory Variable: Variable assumed to cause an effect (e.g., smoking).
• Response Variable: Outcome being measured (e.g., cancer development).

Tips for Identifying Variables

• Consider the logical direction of causation (e.g., Does smoking cause cancer? Does weight affect blood pressure?).
• Focus on what is being measured.

Confounding and Lurking Variables

• Confounding Variable: Variable whose effects cannot be separated from the explanatory variable.
  • Example: Sunlight in a fertilizer experiment.
• Lurking Variable: Variable not accounted for in the study that still affects the response variable.
  • Example: Moisture level in a plant growth study.