The Scientific Process and Experimental Design

Science

approach to explain the natural world based in materialism and empiricism; a community activity and exciting and ongoing; iterative and non-linear process

Materialism

matter; by discoverable, constant rules; nature

Nature

composed of matter & energy; behaves by discoverable, constant rules

Empiricism

relies on evidence obtained through observation & testing

Science is not

based on untestable ideas, able to judge value and/or “meaning”, dogmatic and never changing (static), collection of dry facts in a textbook

Scientific Method

1- observation; 2- ask questions; 3- form a hypothesis; 4- make a prediction; 5- design & conduct experiment; 6- collect & analyze data to draw conclusion; cyclical, not linear

Hypothesis

testable & falsifiable explanation based on what is known to a question

Prediction

If… [restate hypothesis], then… [describe expected comparisons between experimental and control groups]; determines experimental set up

True/False. Most hypotheses lead to one prediction

False

Experiments

test predictions

Sepsis

outcome of an infection that can cause organ damage

Experimental types

manipulative (“controlled”) and natural (“observational”)

Manipulative (“controlled”) experiment

one variable is modified by the experimenter; compare experimental and control groups

Natural (“observational”) experiment

compares differences that already exist between groups; still needs a comparison group

Comparison group

an experimental/manipulated and control/unmanipulated group to compare results

True/False. All manipulative experiments are conducted in labs, and all natural experiments are conducted in the field.

False

Placebo

a pill or treatment that does not contain the substance being tested but is identical in preparation and appearance; ensures that only one variable is changed

Variable

factor that can be set by an experimenter (or in which variation exists and can be studied) to test its effect on a phenomenon

How many variables should distinguish experimental and control groups from one another?

one

Blind Experimental Design

subjects of experiment do not know which treatment they receive

Double-Blind Experimental Design

neither the subjects nor the experimenter do not know which treatment is received

Bias

influencing the outcome of an experiment

Quantitative Analysis

comparing measurements between groups

Statistical analysis

see if differences are likely and not just due to change; Are they significant?

Best Experiments

representative of all the composition you are testing; large sample size; apply mathematical statistics

If data meets the predictions, the hypothesis

is not falsified and is supported

If the data do not meet the predictions, the hypothesis

is falsified and discarded or revised

Probability value

p-value; shows statistically or not differences

Higher than p-value

chance

lower than p-value

statistically signficant

Theory

explanatory hypothesis that has been repeatedly well supported by empirical data