Reliability Task

1.    To collect reliable data of editor's letters in women's and men's magazines, an idea would be to select three women's and three men's magazines and open them up to the page where the editor's letter will be. I Would compare the three men's to each other and then the three women's; picking out linguistic features that are similar in all and the things that all three of the male/female directed magazines have in common. If you were to make a list for men's and women's separately, you would then be able to compare the lists of gathered information and what is similar or different to gather an array of un-biased information.
2.    To collect reliable data of Brian Cox's tweets I would use an idea that was previously stated in the booklet we have been given: selecting every third tweet starting from the most recent. By doing this you would be able to collect un-biased data because by not having a choice of the tweets you would be selecting you are able to identify the lingustic features that are out of your own control so you can analyse whatever you need to.
3.    To collect reliable data of Micheal Gove's comments on education I would watch about five interviews of him on the news, transcribe what is said between him and the anchors in the studio, then try and compare the linguistic features found and apply any theory that would be relevant to the topic I would be researching.