0
$\begingroup$

I'm using interrupted time series analysis to estimate the impact of an intervention in the same group. However, my target variable is a satisfaction index that goes from -1 to 1.

How can I model a dependent variable that is bounded between -1 and 1?

The distribution of the dependent variable looks like this:

enter image description here

$\endgroup$
4
  • 1
    $\begingroup$ Perhaps you could rescale to [0, 1] and use logistic regression? $\endgroup$
    – Scriddie
    Dec 4, 2023 at 22:47
  • $\begingroup$ Please clarify the structure of your data. Do you have a big N, small T panel or just a single time series? $\endgroup$
    – dimitriy
    Dec 4, 2023 at 23:33
  • $\begingroup$ Beta regression is also an option $\endgroup$
    – Alex J
    Dec 5, 2023 at 0:39
  • $\begingroup$ Call this variable $Y$' then $(Y + 1)/2 =: y$ is as @Scriddie advises within $[0, 1]$ and is fit for logistic regression. More at e.g. journals.sagepub.com/doi/10.1177/1536867X0800800212 $\endgroup$
    – Nick Cox
    Dec 5, 2023 at 0:49

1 Answer 1

1
$\begingroup$

You could rescale to $[0,1]$ and then use Beta Regression (Wikipedia) which is designed for response data that lies between $[0,1]$. Also, see this answer to a similar question which talks in a little more detail about the pros and cons of simply using logistic regression for this.

$\endgroup$
4
  • $\begingroup$ Thanks! But how do I interpret the coefficients afterwards? Do I need to scale back? For example, the interruption coefficient (based on a binary variable on whether there was interruption or not) is -0.2133 — does this mean that the index decreases by 0.213 with the intervention? Or do I have to scale the -0.213? $\endgroup$
    – Carolina
    Dec 5, 2023 at 16:18
  • $\begingroup$ Also, how can I perform diagnostic checks on the beta regression? Two of my coefficients (time since intervention and time*intervention) are not significant $\endgroup$
    – Carolina
    Dec 5, 2023 at 16:20
  • $\begingroup$ Perhaps this would help: cran.r-project.org/web/packages/betareg/vignettes/betareg.pdf $\endgroup$
    – blooraven
    Dec 6, 2023 at 3:10
  • $\begingroup$ The seminal resource for beta regression is *A Better Lemon Squeezer& by Smithson and Verkullen. $\endgroup$
    – Peter Flom
    Jan 27 at 22:20

Your Answer

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

Not the answer you're looking for? Browse other questions tagged or ask your own question.