• Question: Can you design a "fair test" to answer your question? In other words, can you change only one factor (variable) at a time, and control other factors that might influence your experiment, so that they do not interfere

    Asked by devan to Bob, Katie, Nisha, Sallie, Vee on 27 Jun 2012.
    • Photo: Sallie Baxendale

      Sallie Baxendale answered on 27 Jun 2012:

      The first thing to do when you are designing a fair test in science is to think very carefully about the question you are asking. The more precise you can be with the question, the better your chance of coming up with a fair test. In experiments that involve people it is often impossible to control for everything because everyone is different, but we try to chosse people who are as similar as possible in the key characteristics that we are interested in when we are studying them.

    • Photo: Vee Mitchell

      Vee Mitchell answered on 27 Jun 2012:

      When I am setting a new experiment we take a lot of trouble to only change one thing at a time. It’s very tempting to change everything at once as you often have lots of ideas about what you want to try to make the experiment work better or be more reliable, but one at a time is the easiest and often the quickest way to get to where you want to be.

    • Photo: Bob Bonwick

      Bob Bonwick answered on 2 Jul 2012:

      yes and no. You can design experiments that look to narrow down variables and then you can make reference ranges (limits or tollerances to the set figures) to your test. you must have a control in your test, both positive and negative, these will show you that something has worked (positive) and also that the occurence of the result isn’t just that of random chance or an accident (Negative). These factors can make the test much more fair, and also alow you to repeoduce your test again and again and get a good base line (standard answer) to draw comparisons on.

Comments