10.1 Scientific Method & Experimental Design
Key Takeaways
- The scientific method follows a logical sequence: observation, question, hypothesis, controlled experimentation, data analysis, and conclusion.
- The independent variable is manipulated by the experimenter, the dependent variable is the measured outcome, and controlled variables are kept constant.
- Positive controls confirm that experimental assays function correctly, while negative controls establish a baseline and rule out false positives.
- Double-blind study designs prevent both participant placebo responses and experimenter bias by concealing group assignments from subjects and researchers.
- Randomization, large sample sizes, and standardized protocols minimize selection and observer bias in scientific research.
The scientific method is a systematic, empirical framework used by researchers, medical professionals, and scientists to investigate natural phenomena, acquire new knowledge, and refine existing theories. On the ATI TEAS 7 Science exam, mastering experimental design is essential. You will be tested on identifying variables, recognizing controls, evaluating study designs, and understanding methods for eliminating experimental bias.
Overview of the Scientific Method
The scientific method relies on structured inquiry rather than intuition or anecdotal observation. It proceeds through a logical sequence of steps:
- Observation: Noticing and describing a specific natural phenomenon, clinical trend, or biological process.
- Question Formulation: Asking a targeted, testable question based on initial observations.
- Hypothesis Generation: Formulating a tentative, testable, and falsifiable statement that predicts the relationship between variables.
- Experimentation: Designing and conducting a controlled study to test the hypothesis under repeatable conditions.
- Data Collection & Analysis: Gathering quantitative and qualitative data and applying statistical analysis to evaluate outcomes.
- Conclusion & Peer Review: Interpreting results to accept, reject, or modify the hypothesis, followed by peer-reviewed publication for scientific validation.
A hypothesis must be both testable and falsifiable. A hypothesis cannot be proven absolute truth; rather, data can support or fail to support it. In clinical research, scientists often construct a null hypothesis ($H_0$), which posits that no significant effect or difference exists between study groups, and an alternative hypothesis ($H_1$), which posits that a specific intervention produces a measurable effect.
Core Variables in Experimental Design
Every well-designed experiment isolates specific factors to determine cause-and-effect relationships. Experiments involve three principal types of variables:
| Variable Type | Definition | Clinical / Biological Example |
|---|---|---|
| Independent Variable | The factor deliberately manipulated or varied by the experimenter to test its effect. | The dosage of a new antihypertensive medication (e.g., 0 mg, 10 mg, 20 mg). |
| Dependent Variable | The factor measured or observed to assess the outcome; changes in response to the independent variable. | The patient's mean systolic blood pressure measured after 8 weeks of treatment. |
| Controlled Variables (Constants) | All external conditions kept strictly identical across all experimental groups to prevent interference. | Patient age range, baseline diet, physical activity level, and duration of drug administration. |
Identifying Variables on the TEAS
On exam questions, identify the independent variable by asking: "What factor did the researcher change or manipulate?" Identify the dependent variable by asking: "What factor did the researcher measure as the outcome?" Controlled variables are any extraneous conditions that are held constant to ensure a fair test.
Experimental Controls: Positive vs. Negative Controls
To ensure that experimental results stem solely from manipulating the independent variable, researchers include control groups. Controls establish baseline expectations and validate experimental procedures.
Positive Controls
A positive control is an experimental treatment group exposed to a variable known to produce an expected positive result. The primary purpose of a positive control is to confirm that the experimental assay, reagents, and equipment are functioning properly.
Clinical Example: In a diagnostic test evaluating a new assay for detecting bacterial infection, a sample containing a known strain of Staphylococcus aureus serves as a positive control. If the positive control fails to yield a positive result, the test reagents are defective, rendering all patient sample results invalid.
Negative Controls
A negative control is an experimental group exposed to all conditions except the independent variable, or treated with an inert substance. It establishes a baseline measurement and confirms that external factors or reagents do not produce a false-positive outcome.
Laboratory Example: In an enzyme activity experiment testing how a novel inhibitor affects lactase, a test tube containing lactase and substrate without the inhibitor serves as a negative control for inhibition, demonstrating maximum expected enzyme activity. In clinical drug trials, a group receiving an inactive sugar pill serves as the negative control.
Blinding and Placebo Controls in Clinical Research
Human clinical trials require specialized experimental designs to prevent psychological influences and investigator bias from distorting data.
The Placebo Effect & Single-Blind Design
The placebo effect occurs when human participants experience measurable psychological or physiological improvements simply because they believe they are receiving an active treatment. To control for this effect, clinical trials utilize a placebo—an inactive substance (such as a saline injection or sugar pill) identical in appearance to the active drug.
In a single-blind study design, the research participants do not know whether they are receiving the active intervention or the placebo, but the researchers conducting the trial know. Single-blind designs minimize participant expectation bias but leave the study vulnerable to experimenter bias during data collection.
Double-Blind Study Design
The double-blind study design is considered the gold standard in clinical research. In a double-blind trial, neither the study participants nor the researchers/clinicians administering treatments and evaluating outcomes know which individuals belong to the experimental or placebo groups. An independent third party maintains the randomization code until data collection finishes.
Double-blind designs eliminate both participant expectation bias and investigator bias (such as subtle differences in how clinicians interact with or evaluate patients in different groups).
Sources of Experimental Bias and Strategies for Mitigation
Experimental bias refers to systematic errors in study design, execution, or data analysis that favor one outcome over another.
Types of Bias
- Selection Bias: Occurs when sample selection does not accurately represent the broader population, such as recruiting only young, healthy college students for a general cardiac health study.
- Observer / Experimenter Bias: Occurs when a researcher's expectations unconsciously influence data collection, measurement interpretation, or interaction with participants.
- Confirmation Bias: The tendency to search for, interpret, and favor data that confirm pre-existing hypotheses while discounting contradictory evidence.
- Sampling Bias: Occurs when non-random sampling methods lead to certain subgroups being systematically overrepresented or underrepresented.
Mitigation Strategies
Researchers minimize bias by implementing randomization (assigning participants to groups purely by chance), using random sampling from diverse populations, increasing sample size ($n$) to reduce statistical noise, standardizing protocols, and employing double-blind procedures.
A clinical researcher tests a new medication designed to lower blood glucose in diabetic patients. Group A receives the new drug, while Group B receives an identical-looking placebo pill. Blood glucose levels are measured daily for 12 weeks. What is the independent variable in this experiment?
During a molecular biology experiment assessing a novel antibiotic, researchers incubate bacterial plates with paper disks soaked in water instead of antibiotic. Why is this negative control step necessary?
Why is a double-blind trial design preferred over a single-blind design in clinical pharmaceutical research?