My Experiment - The Flaws

So now the Lit review/Project proposal has been handed I have gone back to look at my experiment  how it is made up and would like to jot down the main highlighted areas that could go wrong with the experiment and general safety procedures that I wish to carry out to cover myself and others.

First and foremost the film I intend to use is to be an indie film, this is so that no person in the audience can have a preconceived idea as to how to feel towards the main character. It also means I may have full permission from the film maker to use their film in an academic environment. Although I am in talks with a couple film makers to use their films, this does not ensure that they fit the criteria that my music needs. Plan B, if I cannot use an independent film will be to use a lesser known film that has been released. Obviously this means that there is a chance that the audience may have already seen the clips used which is not ideal for the test and also means that I may not have full permission to use the film and will have to be used under strict educational use.

I have written that I wish to use 32 people as subjects as to have a wide pool of results. This may be slightly over ambitious. As a fall back, 24 subjects would be a more manageable number of subjects to use and can easily be broken down into three groups of eight.

All people involved in the subject will be sent a form to tell them exactly what the experiment entails, an example of how the questions will be laid out, will state their anonymity throughout the experiment, state the surroundings that the experiment will take place within along with all times for their attendance. The subjects will be ask to respond to this as an acceptance. The subjects will all be volunteers whether they be peers, students, academics or other. Some backup volunteers should also be required incase any subjects do not attend on the day.

The experiment should take place in a controlled environment, the speakers should not be excessively loud and all in attendance should be comfortable and safe throughout the experiment.

This controlled environment is intended to be a projected screen where the audience can all participate at once. If these plans fall through an online version of the experiment will be available with further instructions as to how to view the video and how to use the questionnaires provided.

Before the experiment begins the subjects will be given the questionnaires for throughout the film and given a briefing as to how to fill them out and what each section means. The subjects will be given a brief explanation as to what the film is but should not  be told that it is about a change in perception as this may sway peoples answers.

THE MAIN PROBLEM 

Now the main problem I have found in my lit review was my quantitative research method. I am right in using a quantitative method of questionnaire that can lead to a statistical analysis of the data received but how I stated to use that data is not the most effective way of gaining accurate results.

The problem lies where I stated that i'd use the mean or average results of data to produce charting evidence of my results. The mean doesn't show a spread of results for example if four people used the rating charts and all though the feeling rated a 1 then the next four people perceived a rating of 7 the average would be 4 which doesn't accurately show the spread of results generated by the audience.

Instead the most accurate way of analysing this data would be through the median, using standard deviation to work out the spread of results and using the interquartile range to further break down the distribution of range.

These methods will be able to show the spread of the audiences results as opposed to one crude average that may not show exactly how the audience responded to the film.