0
$\begingroup$

I have a question regarding the evaluation of my experiment. My study design looks like the following:

Each participant is shown 4 texts, 2 linguistic uncertain and 2 linguistic certain, at the same time. In addition, they are asked how "uncertain" they perceive each of the texts. On the following pages, we asked several questions regarding their socio-demographics, personality traits, and risk tolerance.

The DV is the perceived uncertainty of each text (measured on a 5-point Likert scale). The IVs are the text version (linguistic certain/linguistic uncertain), socio-demographics, risk tolerance, and personality traits.

Examples of my hypotheses are the following:

  1. Age is positively associated with uncertainty perception.
  2. Linguistic uncertainty predicts a higher uncertainty perception.

My problem is: In my resulting data set, I have 4 rows for each participant. Is that a problem for linear regression? Moreover, how do I test my hypotheses? I thought about a hierarchical linear regression to control for other variables when I test for my main hypothesis (2.)

Example of my data:

enter image description here

EDIT: Another DV is the investment sum, the participant would invest in the different companies belonging to the different texts. The possible investment sum ranges from 0 to 10000. Thus, classification is not possible as one DV is continuous. Hypotheses for that DV looks like the following: Linguistic uncertainty predicts a lower investment sum.

Improve this question
$\endgroup$
2
  • $\begingroup$ have you thought about using classification instead of regression? $\endgroup$
    – John
    Commented Oct 5, 2020 at 20:24
  • $\begingroup$ Thank you for your answer. See my edit. I also have a continuous variable so that classification is not possible. $\endgroup$
    – Jensxy
    Commented Oct 5, 2020 at 20:27

0

Reset to default

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.