Skip to contents

A customer study of return intention: whether customers who use a service more come back, and how that use relates to their satisfaction and the offer they prefer. surveyframe gives you the questionnaire, a plan with logistic regressions declared in advance, and the report.

Preview the questionnaire

This is the questionnaire as respondents see it. It is a preview to try, and it stores no answers. Open the Customer satisfaction survey preview in a new tab.

The example analysis plan

Before any data is collected, surveyframe writes the pre-analysis plan: the study, the questionnaire, each research question with its hypotheses, method and decision rule, and the fields a preregistration still asks you to complete. Read the Customer satisfaction survey plan, or download the Customer satisfaction survey instrument (.sframe) to adapt it.

Methods it declares: Binary logistic regression, Ordinal logistic regression, Multinomial logistic regression.

What you get

The study in SurveyStudio, stage by stage: design, collect, check, analyse and report.

The Customer satisfaction survey study on the Design screen, with its questions listed.The Customer satisfaction survey study on the Collect screen, which exports the survey web page.The Customer satisfaction survey study on the Quality screen, which checks the responses.The Customer satisfaction survey study on the Analyse screen, with the declared plan.The Customer satisfaction survey study on the Report screen, which exports the report.

Open this instrument in SurveyStudio

  1. Set up R, RStudio and surveyframe as Get started shows.
  2. To explore the example with its demonstration responses, open RStudio’s Addins menu, choose surveyframe: Try a demo and pick logistic, which opens this template’s study.
  3. To make it your own study, download the instrument above, choose surveyframe: Open analysis workspace from the Addins menu, then Open Instrument in SurveyStudio’s left menu, and select the file.
  4. Adapt the questions on the Design screen and the research questions under Plan, press Save to Studio, then press Download copy on the Report screen to save your study.

Questions

Which model predicts a yes or no answer?

A binary logistic regression, with the odds ratio of each predictor and its confidence interval.

Can I predict a rating instead?

Yes. An ordinal logistic regression predicts an ordered rating, and a multinomial model a choice among 3 or more options.

Can my team use it without writing code?

Yes. SurveyStudio runs the whole study in a window started from RStudio’s Addins menu. R and RStudio need to be installed, as Get started shows.

Optional R code

Open this study in R
d <- sframe_demo("logistic")
render_report(d$instrument)                 # the pre-analysis plan
results <- run_analysis_plan(d$responses, d$instrument)