The purpose of an experiment is to find whether or not two things are related. For instance, does eating too much chocolate affect your weight? Does eating too much chocolate affect your height?
A good experiment should have just one measurable quantity, a variable thing that you can change as an 'input', called the independent variable, and one or more measured things that you suspect change as 'outputs'. All other inputs must remain constant. For instance, you may set up an experiment where lots of girls and boys of the same weight eat varying amounts of chocolate over a month (the input), and measure their weight and height after a month (outputs).
After the experiment, you look at the changes in the inputs, and look for a corresponding change in the outputs - more chocolate leads to a growth in height (unlikely) or more chocolate leads to an increase in weight (likely). You analyse the data statistically, to see how likely it is to get the outputs from the inputs by chance. The NULL HYPOTHESIS is that the inputs and outputs are unrelated - that any apparent effect is random. You hope do DISPROVE this hypothesis statistically, and show that it is extremely unlikely that the pattern is random. You can then draw the conclusion, the ALTERNATIVE HYPOTHESIS, that it is more credible that the inputs and outputs are causally linked than that the results you see are simply due to chance. You may then need to create further tests to identify the mechanism by which the inputs affect the outputs.
It is worth noting that most scientific experiments fail: the null hypothesis is not disproved.What is the 'independent variable' in an experiment?
The independant variable is the thing you change in the experiment, the dependant is the thing which you measure, and im not entirely sure about the null hypothesis. :)What is the 'independent variable' in an experiment?
Independent variable is the one you change, dependent is the one you measure - it is your result, im not sure about hypothesis sorry
No comments:
Post a Comment