Saturday 18 October 2014

Survey link

Link to survey: https://www.surveymonkey.com/s/RQPPGC3

I can now begin to collect data.



Designing response surey

To see the aspects of what went well in my film, as well as what I could improve, I am designing a survey which can be completed online. The viewer will find a link to the survey after watching the film.

Check list for designing a survey:

1. Define objectives - what do I want to find out? What decisions am I trying to make?
2. Work backwards - Once objectives are set, determine the data needed to make decision.
3.Check for bias - Ensure there are no leading questions.
4. Test run - with small sample size. Check clarity and logic.
5. Collect results and analyse data - make sure it has been sent to correct target audience.

1. My objectives: Main - Have I answered my question? Does the viewer have a better understanding of laser eye surgery after watching the film?
Others - Are there any unanswered questions still? Any addition info they wish they had?

2. Data needed: I would like to have a large age range of target audience therefore need to know age and sex of viewer. Need to know at what stage have they considered laser eye. How much they knew about laser eye before vs. how much they know after. Additional info such as ways to improve.

3. All multiple choice questions will have an array of different choice of answers.

4. I will send a test survey to close family and friends as a prelim. It will be a quick way to check everything is smooth.

5. To send my survey to...
<18's - I will send the link via social media sites, such as Facebook and twitter.
For 18-50 year olds - I will also use Twitter, as well as emailing the link to many people.
For 50+ year olds - I will either email or ask people who are at Moorfields eye hospital to complete it directly after watching the film then and there.

To obtain a large sample size, I will go beyond just showing the film and survey to people with an interest in having eye surgery. A large sample size ensures the mean is reliable as extreme values and answered do not have as big an effect.