Personal data from Internet users is a sought-after commodity today. The gathering, accumulation and storage of this data can have various negative consequences for users. While users, upon request, are concerned about the security of their data, social networking usage or online shopping sales figures raise doubts. The phenomenon of inconsistency between respondents' statements and their privacy behavior is called privacy paradox in the literature. Therefore, it is necessary for research in this context to record actual behavior in order to be able to make durable statements regarding influencing factors and potential interventions or support. Here, the difficulty of investigating the behavior of users in a risky and as realistic a situation as possible, without exposing them to an actual risk, must be addressed.
The initial aim of the work is therefore to identify a suitable possibility to empirically record actual privacy behavior in online shopping. Further aims are the translation of the recorded behavior into relevant data and the determination of significant predictors.
The first step is to develop a definition for the necessary operationalisation of the construct of actual data protection behavior in online shopping. With the help of an expert survey, the eight most important hints which users can use on the website of an online shop to gain an impression of the respective handling of personal data are identified. In order to be able to quantify the extent to which this information is used within an eye tracking analysis, the measure of fixation is defined within the framework of this work.
The developed study design is the basis for the evaluation of a working model regarding potential influencing factors on actual behavior in online shopping. An explorative study leads to first findings regarding possible predictors. To validate these findings, the working model is adapted and a validation study is carried out, the design of which corresponds to that of the exploration study.
The analysis of the eye-tracking patterns of both studies shows that very little use is made of all of the indications. The shop name (URL) is fixated most frequently (23 of 73 participants). The highest value for the variable of actual behavior in online shopping, which is the total amount of the most important indicators considered by the user, is only four out of eight.
The age of the participants proved to be a significant predictor, which, however, only explains a small proportion of variance. The insignificant correlation between the probability of surveyed behavior and the corresponding actual behavior in this study is in line with the findings on the privacy paradox. This proves that intention is not an accurate predictor of actual behavior in the context of data protection and reinforces the necessity of empirically recording actual behavior for further research in this context. The research concept developed in the course of this work can provide a framework and be used directly for further research, which includes actual data protection behavior in online shopping as a variable. |