• Question: what kind of controls did you use to conduct your research?

    Asked by Emily Key to Ben, Sam, Kirsty, Maggi, Rose on 13 Jun 2017.
    • Photo: Kirsty Miller

      Kirsty Miller answered on 13 Jun 2017:

      Hi 🙂 If you mean by controls, the things we did to try to get the most accurate results, I had a few main ones. I collected data from several schools in different towns to try get a mix of students from different backgrounds (this was to get what we can a ‘representative sample’ – or a sample of participants that represents, or is similar to, the normal population).
      Because I collected questionnaire data I made sure the questionnaires were anonymous to try to get as accurate results as possible. Often people can be worried to give truthful answers if they feel they may be judged on them (or get in trouble for them!) so I made sure students knew that there was no way that I could match their answers to them. They got unmarked envelopes to put their questionnaires in, they completed them under ‘exam conditions’ and then put their envelopes in a big box. This made it clear to them that no-one would know what they had said. I got some quite personal responses so I think the pupils trusted that their answers were private!

    • Photo: Rose Turner

      Rose Turner answered on 17 Jun 2017:

      Hi Emily, I tend to work with adults form the same population – in other words, I don’t use a particular clinical or developmental group and a separate group of ‘normal’ controls. However in my experiments I still need to include control variables. I have been doing lots of research around how reading stories affects people’s performance on different tasks so for that kind of research I would give one group a story, and one group something else to read but not a story (something like a description of how something works – this is called ‘expository nonfiction’). The difficulty comes in matching the two texts as much as possible (if they are too different in terms of word count and difficulty, that can cause a problem). I might also include a ‘no reading’ group so that I can check that any findings ween’t simply to do with the act of reading, but were to do with reading the fictional story, specifically.

Comments