to Pretest Database Pretest Database
Project Title:German Longitudinal Election Study (GLES) 2021 (English Version)
  1. General Information: *Note: The item was tested in German. This is an English translation of the original German wording.
  2. Question Text: On how many days in the past week did you see or read posts about political issues on social networks, such as Facebook or Twitter?
    [An wie vielen Tagen in der vergangenen Woche haben Sie in sozialen Netzwerken, wie z. B. Facebook oder Twitter, Beiträge über politische Themen gesehen oder gelesen?]
  3. Answer Categories on ... days [an ... Tagen]

    Not seen or read in the last week [nicht in der vergangenen Woche gesehen oder gelesen]

    1. Recommendations: Question: We recommend rewording the question to ask for the past seven days, as this time period is clearer and may make it easier to answer the question.

      Instruction: We also recommend adding an instruction on the format of the answer: "Please enter a number between 0 and 7."

      Response format: If the fill-in instruction is added, the "not seen or read in the past week" re-sponse option should be deleted. However, if the response option is to be re-tained, which we strongly discourage, it should be reworded:
      "not seen or read any posts about political issues in the past seven days."
  1. Cognitive Techniques:Information image/link to cognitive pretesting General Probing, Difficulty Probing, Comprehension Probing, Specific Probing
  2. Findings for Question: 1. How do respondents go about answering the question?
    Most respondents (n = 8) said that they first considered how often they normally visit social network sites and then how often they encounter political content. They had not explicitly thought about the time period "in the past week" mentioned in the question, but this did not seem to affect the accuracy of their answers due to the regularity with which the persons used social networks:
    • "I mainly thought about whether there was a difference between weekends and during the week, and then realized that there are workdays when you don't have much time for that. [...] And then I tried to bring in an average of how many days I use social media during the week and what comes my way about political topics. I was thinking that there's definitely one or two days during the week that I'm too busy to consciously notice that I've read or seen anything political." (TP 03, response: 5 days)
    • "I first thought about how often I open [social networks], which network I use. [...] Then I thought about how often I open that. [...] By knowing my feed, I know that I read or see political topics every day. That's why I entered 'on seven days'." (TP 04, response: 7 days)
    • "I was thinking that I'm on social media seven days a week and that automatically seven days a week politics is shown." (TP 09, response: 7 days)
    Only persons 01 and 10 explicitly thought about their behavior in the past week when answering the question:

    "Now I have briefly thought about what was so extreme in the last week, and there was just the Corona topic again very high up. [...] I'm only starting from Facebook now, because I'm only there. As far as something like that is concerned, you've seen a lot of reports about Corona and the whole best ends." (TP 01, response: 4 days)

    2. How difficult is it for the respondents to answer the question?
    With the exception of subject 03, all persons found the question to be "very easy" (n = 5) or "rather easy" (n = 4) to answer, especially based on the regularity with which they used social networks (or did not use them at all):
    • "Because I know I don't [consume] political issues through these networks or social media." (TP 05)
    • "Because, like I said, I'm shown something every day and I'm on Facebook every day." (TP 02)
    Respondent 03 found it "rather difficult" to answer the question due to the specification of an exact number of days: "That I should specify the exact number of days, I find that difficult."

    3. How do the respondents interpret the instruction "in the past week"?
    Slightly more than half of the persons (n = 6) interpreted the period "in the past week" as the span of the past week from Monday to Sunday:
    "The past week from Monday to Sunday." (TP 07)
    In contrast, four persons (TP 02, 04, 06, 09) understood it to mean the past seven days:
    "From today one week back, that is, until last Friday." (TP 02)
    The phrase "in the past seven days" was judged by the respondents to be more precise and thus easier to answer:
    • "Yes it would, it would make a difference because maybe last week was a bit longer ago and I can't remember it as clearly as the last seven days, depending on when I answer the question-naire. If I answer it now on Sunday, it could also mean the week before." (TP 03)
    • “In the past week' is another term for me. With 'in the last seven days,' I would feel more like it was actually specifically asking about the last seven days. With 'in the past week' I just feel like it's supposed to help me make some estimate, but theoretically it wouldn't matter that much when those seven days occurred." (TP 04)
    4. Other noticeable problems
    Respondents who have not seen or read any posts about political issues on social networks in the past week can answer the question in two ways. First, they can insert a zero in the open response field (like TP 07) or check the response option "not seen or read in the past week" (like TP 05). To facilitate data preparation and analysis, this could be standardized, e.g., by adding a fill-in instruction instead of the additional answer option: "Please enter a number between 0 and 7."
  1. Question Topic: Digitalization/ Use of digital systems
  2. Construct: Frequency of political content when using social media