Challenge Definition

During a student project, the city administration of Lucerne invited city residents to answer an online questionnaire about their perceived quality of life. The result of this online questionnaire is a rich set of data from 630 participants.

Already during the student-project, many interesting insights could be read from the data. Nevertheless, we would like to dig a little deeper in this set of data.

• What coherences can be found?
• How do the demographics influence the different aspects of life that are connected to quality of life?
• What is your advice to the city administration?

The city administration is happy to use the results of this challenge to further improve quality of life in the city of Lucerne.

Data:

• Answers of the online questionnaire (raw data, unpublished)
• Results of the student project
• Results of the (official) population survey 2019: www.lustat.ch/daten?id=26999

Contact: nadja.vonballmoos AT stadtluzern.ch

Project

Our Approach

To get to our overall goals we have different approaches.

• We build clusters to group the data.
• We look at correlations and use machine learning models (LM, Decision Tree …) to see which factors impact the overall quality of life the most.
• We use text analysis to interpret the comment sections of the online questionnaire.
• We search for specific demographic characteristics with a high or low quality of life to get further insights.

Main Results and Insights

Disclaimer: The data in this survey is not representative of the whole population of Lucerne

Clustering
By appling clustering on the whole data set, we could find 5 groups that differ in various aspects of their quality perception. Correlation Matrix
The Correlation Matrix helps us to understand, which variables correlate to a certain degree with the response variable "Lebensqualität Allg."

Decisicion Trees
The Decision Tree gives us a idea, which variables could be relevant for the splits.



Closer Look to City Participation
With help of logistic regression and classification trees we found that especially the variables "Zuzug" (time since moving in to Lucerne) and "Bildung" (eduction) have an influence on the city participation and engagement.



Text analysis
A text analysis of the answers to the question "Wovon wünschen Sie sich mehr?" led to the following word cloud.

Next steps

• Further focus to more specific questions (based of the new insights)
• Further analysis with more survey data
• Build a framework including the most common used methods