Consulting

The Seminar for Statistics (SfS) at the ETH Zurich offers a statistical consulting service, contractual data analysis, as well as statistics and software courses.  

Many scientific studies are based on the interpretation of measurements or observations, thereby including statistical analysis. The Seminar for Statistics offers advice on experimental design, the choice of the stastistical method, the choice of an appropriate computer software for your statistical problem, as well as help for the interpretation of the results. Please contact us already while designing your experiments or field study.

General information

Usually, this will take place in a single one-hour meeting with one or a few follow-ups if needed. The goal of our meeting is to give you some guidance on how you can perform an appropriate statistical analysis on your own. We have a broad knowledge in statistics but we are not necessarily experts in your field. Moreover, we don't have the capacity to do the analysis for you. Think of us as statisticians who can give you a second opinion, discuss the problem with you, give you some hints, and point you in a direction of further reading. A large part of our first meeting consists of understanding your study. This part is very important for choosing an appropriate statistical method.

For members of ETH-Domain, this service is free of charge. Appointments for statistical consulting can be arranged with the consulting team, using the contact form with an attached short description of the study (Downloadsee here for an example (PDF, 105 KB)).

Analysis of data (for ETH-institutes and external companies) can be carried out in terms of a project. After a cost estimate the collaboration will be contractually regulated. In most of the cases the minimum effort for processing a project equals a working week. A detailed report can be provided, if requested.

Examples of topics of concluded studies:

  • Measurements of radiation have been collected from 40 samples of a chemically polluted site. How can the average radiation for the whole site be estimated, including a precision of this estimate?
  • A device for measuring blood sugar continuously, without taking blood samples, is under development. It is attached to the skin and measures electrical impedance signals as well as temperature and optical signals. The challenge is to develop a model that relates these signals to blood sugar and allows for predicting the latter with sufficient precision.
  • Manufactoring plastic material used for bottles is protected by a patent if a certain property exceeds a given threshold. A sample is used for verifying whether this is the case. Depending on the way of summarizing the data, the conclusion ranges for "clearly violated" to "clearly not violated". Which analysis is adequate?

Prospective clients should contact the consulting team for a non-committal discussion.

Unlock the Potential of Your Data with Expert Assistance - For Free!

Are you facing a data analysis challenge and could use some expert help at no cost? Our Statistical Consulting Service is seeking projects for our applied student seminar within the Master of Statistics program.

What Benefits Can You Expect?
- Complimentary Data Analysis: Your data will be analyzed by skilled students from the MSc Statistics program, under the guidance of the Seminar of Statistics, D-MATH faculty.
- Expert Recommendations: Gain insights on methodologies and strategies best suited for your data.
- Valuable Resources: Receive the R code used for analysis, complete with detailed explanations.

Project Criteria:
- Timing: Projects should have flexible deadlines; results will be ready by early June.
- Data Availability: Your data should be ready by Feb/Mar. Planning for an experiment? No prior data needed.
- Effort Guarantee: While success can't be guaranteed, our students dedicate considerable effort to each project.
- Short-term Engagement: Detailed queries will be addressed in June, followed by a handover of all materials. Future support is not included, allowing you to adapt the students' work for your continued analysis.
- Data Sensitivity: We cannot accept projects with highly confidential data.

How to Get Involved:
- Initial Contact: Drop me an email with a brief description of your project.
- Engagement: Dedicate about 1 hour in Feb/Mar to outline the problem to the students and be available for occasional follow-up queries through June.
- Final Meeting: Spend roughly 1 hour in June to review the findings and receive all materials.

Timeline:
- By Mid of January: Express your interest to participate & send description.
- End of February: First meeting with students to discuss the project and data.
- February to June: Ongoing project work, with occasional communication.
- Beginning of June: Receive a comprehensive report, final presentation, and all relevant R-Code.

Get Your Project on the List!
Interested? Send a concise description (about five sentences) of your data and study objectives to . Your project will be featured in the list from which students select their projects. Make it engaging to capture their interest! If you have multiple research questions, please prioritize them due to the time constraints of the semester.

external pageR is GNU software, and can thus be downloaded free of charge via the Internet, for instance from the ETH mirror of the external pageCRAN network.

Consulting

The Seminar for Statistics offers consulting services for R. Since R is the main statistical software used by the Seminar for Statistics, our consulting services are at their best when this software is involved.

Courses

On we also offer introductory and intermediate courses for companies, for instance in-house courses.

external pageZurich R Courses offer a series of introductory and advanced courses on statistical data analysis with R on a regular basis.

Relevant lectures

The following lectures are focused on using R:

  • Using R for Statistical Data Analysis and Graphics.
  • Applied Statistical Regression. Its tutorials/exercise hours make intense use of R.
  • Applied Analysis of Variance and Experimental Design.
  • Computational Statistics. The exercises are heavily focused on using R.

An introductory class in statistics is a prerequisite for the last three lectures listed above (see course catalogue for more details).

Hints

User Group

There are mailing lists for R users. For more information and to subscribe, see external pageThe R Project for Statistical Computing (Mailing Lists). The primary mailing list is called "R-help"; it offers swift and competent answers to problems with R.

Newsletter

Since January 2001, R has had an online external pagenewsletter, which in 2009 became the external pageR Journal.

Contact

  • +41 44 632 3505

ETH Zurich
Seminar für Statistik (SfS)
Rämistrasse 101
HG G13
8092 Zürich
Switzerland

JavaScript has been disabled in your browser