Week 17 Flashcards
What is science?
Knowledge or a system of knowledge covering general truths or the operation of general laws
Obtained and tested through the scientific method
What is the difference between testable and non-testable questions?
Testable - answered with specific activities
Non-testable - don’t have a definitive answer
E.g. testable: is the sky more blue on a clear summer than on a winter day
non-testable: why is the sky blue
What is a hypothesis?
Statement about the world, whose truth is being tested
What properties must a hypothesis have?
Precise and testable
Falsifiable - aka you can prove it’s wrong
What is wrong with the following hypotheses:
“If electrons were 10% less massive, no life would exist “
“Macs are better than PCs”
Not testable - can’t test against all life forms
Not falsifiable - a PC enthusiast would always be able to disagree
Who said hypotheses must be falsifiable, and when?
Karl Popper, 1959, in “The Logic of Scientific Discovery”
What are the 3 steps involved in conducting hypothesis-driven research?
- Frame a hypothesis
- Design and conduct experiments to test it
- See if correct or disproved
When is a hypothesis appropriate or inappropriate, and why?
Appropriate for small number of cases to decide between, e.g. which of a few drugs is better or something
Not good for vast number of possibilities, such as “The density of this rock is 0.1g/cm3”
“Measuring a Value” is a type of ___________ _________
Scientific research
“Measuring a function or relationship” is a type of _________ ___________
Scientific research
How is a function or relationship measured?
Control one variable (independent), measure the response in another (dependent), keep all the rest constant (actually called control variables)
Often, measured functions/relationships can be linked to an original _________
hypothesis
What is “constructing a model”?
Type of scientific research
What is the purpose of constructing a model?
Can be used to predict outcomes
Can form the basis of new models which extend the original
“Predicting London’s weather for next week” is an example of what type of scientific research?
Constructing a model
What is observational or exploratory research?
Observing or testing things without any particular hypothesis in mind, to see if any patterns emerge
What is industrial or applied research?
Research done in response to market/demand for some type or product or facility
e.g. aircraft design and testing
What are 6 types of research method in science?
Hypothesis driven research
Measuring a value
Measuring a function/relationship
Constructing a model
Observational/exploratory research
Industrial/applied research
What steps are involved in hypothesis-driven research?
Identify independent and dependent variables
Frame null and alternative hypotheses
Collect data
Analyse data
Accept or reject null hypothesis
What will the alternative hypothesis always say?
Changes in independent variable lead to changes in dependent variable
What is a Type I error in hypothesis testing?
Researched rejects the null hypothesis even though it’s actually correct (false positive)
Aka incorrectly predicting a change in independent leads to change in dependent
What is a Type II error in hypothesis testing?
Researcher accepts null hypothesis even though it’s false (false negative)
aka ignores a real effect of independent variable
What is bias?
Actions that lower the validity of an experiment, willfully or otherwise
What are some examples of bias in experiments?
Doctor knowing which group gets a placebo
In data-driven tests, collecting only data which fits the predicted model
What errors can happen when collecting measurements? How can they be ameliorated?
Random error/noise - collecting lots of data smooths this out
Systematic error - some consistent distortion of measurements. Ensure all equipment is working correctly, calibrate it