Experimental design

[Stick]

I'm not very good at these types of questions. I'm assuming the theory behind this was covered in Biology Units 1&2, which I didn't do. Would anyone mind giving me a brief crash course on this? It's obviously something that's going to pop up fairly frequently.

[ealam2]

Experimental Design

-only allow one condition to vary so if anything else in your experiment changes you can conclude that the condition that you varied (independent variable) caused this change
 
Things that can change in an experiment are called variables.

Independent variable :  variable being changed in the experiment
 
Dependent variable: variable that you measure, that you expect will respond to changes you have made to the independent variable

Controlled variables : All the other variables are kept the same in each trial so that is a fair test.

Controls: A control group is one in which all the variables are controlled but the independent variable is absent, to prove that only the independent variable will have the effect. It is a reference by which the other tests are compared.

[Yacoubb]

Also make sure when you use organisms, use quite a large number (e.g. 100 divided into two groups in your population), and all organisms are of same species, age, gender, health state, nutrition state, diet, temperature, water supply, to minimise any extraneous variables and control variables to increase reliability of your findings.

Repeats - repeating the experiment decreases the risk of systematic or random error and further strengthens the reliability of your findings that support or negate your proposed hypothesis.

[ealam2]

As Yacoubb said, repeating an experiment is vital. (ideally minimum of ten times but that's not possible when doing pracs at school)

Also in any experiment, you should include the safety and errors or limitations of the experiment.

[Yacoubb]

Also in any experiment, you should include the safety and errors or limitations of the experiment.

In an exam, you'd only need to write the Hypothesis, Method and Findings that would either support/negate your hypothesis. Safety is not really necessary. Limitations would practically be asked in a separate question, however by stating that they are of all same age, sex, health, etc, you are implying that you are in fact controlling variables to avoid random or systematic errors that can distort your findings.

[forchina]

are you also meant to discuss observations that you couldn't see but knew occured in the analysis section. even if it asks you what processes are taking place?