Findings Web Probing:
Respondents were randomly assigned to one of the three question versions with clarifications shown on demand (ToolTip), to all respondents (via instruction) or without clarification. In total, 62% (n = 491) of respondents reported that they were employed, 38% (n = 300) reported to be self-employed, and only one respondent answered question Q7 with “Don’t know” (see Table 14). There were no significant differences between question versions (χ2(4,792) = 3.022, p = .554) or countries (χ2(4,792) = 3.236, p = .519).
Based on the response distributions, there is no indication whether or in which way clarifications are presented impacts response behaviour. Cognitive interviews were used to gain insights on possible silent misunderstandings, that is an incorrect understanding and subsequent self-assignment of respondents to the answer categories.
Findings Cognitive Interviews:
In the cognitive interviews, respondents were first shown question version 1 (ToolTip) and after discussing this question version, they were shown question version 2 (instructions displayed directly) on a separate screen.
In Germany, nine respondents declared themselves as employees and seven respondents as self-employed. In Poland eight respondents declared themselves as employees and seven respondents as self-employed. Additionally, one respondent (PL15) could not assign himself to one of the options.
While in Germany there were mainly no problems with the understanding of the two terms “employee” and “self-employed”, in Poland different understandings of “employee” were mentioned. This is because in Poland the term for “employee” (PL: Pracownik najemn) can have negative connotations. Therefore, a question was explicitly asked in Poland whether the term “employee” was actually perceived negatively.
Eight Polish respondents stated that the term "employee" had negative connotations for them. They mostly understood it to refer to working arrangements based on another type of contract than employment contract (of mandate, specific work) or simple jobs such as being hired, without any social security or rights:
This assessment most likely stems from the characteristics of the Polish labor market with very complex regulations which do not protect the weakest players of the labor market.
The other eight Polish respondents defined “employee” in neutral terms, as a person employed based on any kind of contract or working for someone else’s business, also as a subcontractor.
Apart from the negative connotation of the term “employee” in Poland, two respondents (PL07, PL15) who owned a company with limited liability had problems classifying themselves correctly:
Additionally, in Germany two respondents (DE03, DE05) had problems understanding what was meant by “family workers”:
Except for two respondents (DE09, PL04), no one used the ToolTips to answer this question. Most of the participants explained that they knew the answer to this question without looking up clarifications and that it was easy for them to differentiate between employee and self-employed:
In addition, seven respondents (DE: n = 1, PL: n = 6) explained that they did not use the ToolTips because they either did not know they could use it, did not notice it, or forgot about it:
In both countries, the explanations of the terms “employee” and “self-employed” in the ToolTips were found to be informative. In Poland, seven respondents preferred the version with the ToolTips, eight preferred the version with the instructions shown directly and one respondent (PL02) found both versions equally good. In Germany, eleven respondents preferred the version with ToolTips and five the version with the instructions shown directly.
Respondents preferring ToolTips stated that it is easier to answer the question in this format and that it is clearly arranged:
Respondents preferring the instruction shown directly explained that they would probably oversee the ToolTips, because they are hardly intuitive:
One respondent (PL06) explained that there is no consistency in how the information is displayed in the questionnaire, that’s why he overlooked the ToolTips:
Summary