Observational Studies, Designed Experiments, and Confounding Variables

Explanatory and Response Variables:
- Explanatory Variable: A variable that is manipulated or categorized to explain, predict, or observe changes in an outcome.
- Response Variable: An outcome variable whose value depends upon or responds to changes in the explanatory variable.
Observational Study:
- Definition: An observational study measures the value of the response variable without attempting to influence or manipulate the response or explanatory variables.
- Core Characteristics: Researchers passively observe and record data on subjects without subjecting them to experimental interventions, treatments, or assigned groups.
Designed Experiment:
- Definition: A study is a designed experiment if a researcher randomly assigns the individuals in a study to specific groups, intentionally manipulates the value of an explanatory variable, and controls other explanatory variables at fixed values.
- Core Characteristics: Involves direct influence, intentional manipulation of treatment variables, and strict experimental controls over confounding factors.

Example 1: Mobile Phone Usage and Brain Tumors (Observational Study):
- Objective: Determine whether there is an association between mobile phone usage and the occurrence of brain tumors.
- Explanatory Variable: Mobile phone use (measured by how frequently or how long subjects use smartphones).
- Response Variable: Presence of a brain tumor (whether or not an individual develops a tumor).
- Classification: Observational study. Researchers simply observe women over time without controlling phone usage habits, dictating exposure lengths, or applying experimental treatments.
Example 2: Radiation Exposure in Rats (Designed Experiment):
- Objective: Evaluate the biological effects of cellular radiation on tumor formation using animal models.
- Experimental Structure:
- Group 1 (Control/Baseline): Rats receive no radiation exposure ( $0$ exposure level).
- Group 2: Rats receive GSM (Global System for Mobile Communications) radiation.
- Group 3: Rats receive CDMA (Code Division Multiple Access) radiation.
- Explanatory Variable: The specific type and amount of radiation received (No radiation, GSM radiation, or CDMA radiation).
- Response Variable: Brain tumor development (whether or not the rats develop tumors).
- Classification: Designed experiment. The experimenters intentionally manipulate the radiation type and force random assignment into strict treatment groups.

Study Details:
- Target Population: Seniors aged $65$ and older.
- Sample Size and Duration: Records of over $36,000$ seniors observed over a $10$ -year period.
- Group Division:
- Group 1: Seniors who chose to receive a flu vaccination shot.
- Group 2: Seniors who chose not to receive a flu vaccination shot.
Observed Results:
- Seniors who received flu shots were $27\%$ less likely to be hospitalized for pneumonia or influenza.
- Seniors who received flu shots were $48\%$ less likely to die from pneumonia or influenza.
Study Design Classification:
- Classified as an observational study because subjects self-selected whether to receive the vaccine rather than being randomly assigned to treatment and control groups by researchers.
- Explanatory Variable: Vaccination status (whether or not the senior received a flu shot).
- Response Variable: Health outcome (hospitalization status or mortality due to pneumonia/flu).

Limitations of Definitive Causal Claims in Observational Data:
- Observational studies cannot definitively establish direct cause-and-effect conclusions because underlying external factors may drive observed differences between groups.
Lurking Variables:
- Definition: A lurking variable is an explanatory variable that was not considered or measured in the study, but that directly affects the response variable.
- Examples in Senior Health Studies:
- Immunocompromised Status: Baseline immune health differences heavily influence mortality risk regardless of vaccination.
- Pre-existing Health Conditions: Co-morbidities affecting susceptibility to severe flu complications.
- Specific Age Differences: Health risks vary significantly across age brackets (e.g., comparing an individual aged $65$ to an individual aged $80$ ).
- Health Consciousness: Individuals who actively seek out flu vaccines may engage in other beneficial health habits (nutrition, exercise, hygiene) compared to those who decline vaccines.
- Community and Living Conditions: Environment, living facility type, and social exposure levels.

Confounding Variables:
- Definition: A confounding variable is a variable in a study that affects the response variable and whose effect cannot be distinguished from the effect of an explanatory variable.
- A confounding variable may or may not be associated with another explanatory variable included in the study, making it impossible to isolate which variable causes the change in outcome.
Example: Online Homework vs. Textbook Homework Study:
- Research Objective: A professor wants to determine whether online homework or traditional textbook homework yields better student performance.
- Experimental Setup:
- Morning Class: Assigned online homework.
- Evening Class: Assigned traditional textbook homework.
- Explanatory Variables: Homework format (Online vs. Textbook) and Class Time (Morning vs. Evening).
- Response Variable: Student success (measured by final exam scores).
- Confounding Issue:
- Class time (Morning vs. Evening) is confounded with homework format.
- Student characteristics may naturally differ by time of day (e.g., morning students may possess higher natural motivation or alertness than evening students).
- Because homework type and class schedule vary simultaneously, the professor cannot determine whether score differences are caused by the homework format or the time of day.