In my previous blog post I promised this one will finally have pictures or at least some graphs and will not be text-only again. Unfortunately, there are no results to showcase yet. [I actually wanted to include a GIF of Kevin from "Home Alone" saying "I'm sorry" at this point, but it seems GIFs are not properly supported by the website. I guess I am sorry twice now.] However, this provides me with the opportunity to explain why it is so important to take one's time to creating meaningful surveys. 

The importance of a meaningful survey

Brace noted in his book about questionnaire design for market research [Bra18], even the best questionnaire relies on people and what they (can / do not) tell us. Brace also looked at the impact of wording, response categories as well as layout of the questionnaire on the final result. To avoid major problems, he suggests to at least run an informal pilot of the questionnaire. Even though the survey for my master thesis will not focus on the field of market research, most of Brace’s findings are also applicable.

The survey is, overall, the most crucial part of my thesis, its results will be the basis on which I select certain user tracking tools/methods for a more furrow legal analysis. I already anticipated creating and distributing my questionnaire will yield the most problems.

My plan, as stated in my proposal, was the following: "To answer the second question [Annot.: Which (state-of-the-art) user tracking technologies are currently used by Austrian enterprises for business purposes?], an online questionnaire will be developed based on the outcomes of the first research question [Annot.: What is currently considered state-of-the-art user tracking and what other methods are still in use?]. The questionnaire will be distributed to various Austrian enterprises to ensure a big enough sample size."

As I already mentioned in my last blog post, while answering the first research question I realised I needed to take my research to a less abstract level and match user tracking methods with actual user tracking tools. This way I hopefully bridged the gap between expert-only knowledge and consumer-level knowledge and gave me the chance to create an actually answerable survey. I also tried to create a short and precise survey, knowing that the number of answers decline the longer a questionnaire gets [Deu04].

The second problem I encountered was questionnaire distribution. There is currently no easy nor good way to reach a representative sample of Austrian enterprises. In a worst case scenario, I have to reduce my target group to Austrian enterprises with a web presence. This however will most definitely leave me with biased results.

Literature

[Bra18] Ian Brace.Questionnaire design: How to plan, structure and write survey material for effective market research. Kogan Page Publishers, 2018.

[Deu04] Deutskens, E., de Ruyter, K., Wetzels, M. et al. Response Rate and Response Quality of Internet-Based Surveys: An Experimental Study. Marketing Letters 15, 21–36 (2004). https://doi.org/10.1023/B:MARK.0000021968.86465.00