Before you analyse and evaluate the data of your survey, you should formulate your expectations and guesses concerning the results – this is called generating hypotheses. You can use these hypotheses later on when evaluating your data to test your hypotheses: Can you verify or falsify them?
Samples for possible hypotheses:
A) Hypotheses concerning simple frequency "The majority of interviewees feels insufficiently informed about the EU.” This hypothesis suggests that more than half (50%) of the students answered “no” to the question 13 “In general, do you feel sufficiently informed about the EU?” B) Hypotheses concerning contexts that are more complex “The more informed interviewees feel about the EU, the more likely they are to have a positive image of the EU.“ This hypothesis implies a correlation between:
Feature 1: to feel informed about the EU (question 13)
Feature 2: to have a positive image of the EU (question6)
Please note: Hypotheses implying correlations often include words like “rather than“, „compared to“, or the like. Such hypotheses are usually tested by means of contingency tables.
Read the questions on the questionnaire about Europe (M02.05) again carefully.
Work in pairs and generate four hypotheses which can be tested by means of the questions your group works with. You can include questions that are tackled by other groups, in case you want to generate hypotheses implying correlations. Note down the features (questions) that are supposed to be analysed when writing down your hypotheses (hypotheses concerning simple frequencies = one feature; hypotheses implying a correlation = two features).
Optional assignment for fast working groups: Are there hypotheses you are interested in which cannot, however, be analysed by means of the survey at hand because the relevant questions are missing? Which questions would have had to be included to make testing your hypotheses possible?
1. Hypothesis (simple)
This hypothesis concerns the results of feature:
________________________________________________ (question: ___)
2. Hypothesis (complex)
This hypothesis implies a correlation between
feature 1: ___________________________________________________ (question: ___) and
feature 2: ____________________________________________________question:___)