Terms of service
rules of engagement
The ICDS acts for its statistical service in accordance with the ethical code established by the American Statistical Society.
Consultants adhere to the principles of professional and scientific integrity, clients (researchers) are expected to do so as well.
Consultants aim to offer the most appropriate scientific solution to successfully perform the study, obtain statistics, understand and communicate results.
Consultants refrain from performing any analysis when encountering problems with the study design, hypothesis, research question or data collection.
A client is expected try to specify the project, experiment, hypothesis, research questions, as clearly as possible, including all marginal conditions. This typically includes a layman introduction of the nature of the data and the data collection, with a focus on what is statistically relevant. The client is expected to signal possible problems with the data collection and measurement process. Any anomaly or deviation from protocol should be pointed out by the client, with reference to possible causes and remidiating actions.
The client is expected to put in the necessary effort to understand the information offered by the consultant, and to follow up on suggested study material.
The client is expected to translate the statistical results to their own field of expertise. The client will, together with the consultant, discuss the correct interpretation of results.
Statistical and methodological support is offered to all researchers of the Life Sciences and Humanities within the VUB / UZ Brussel community.
Consultancy can be provided to all members of our community in response to a client's request.
Support is offered on the design of studies - sample size justification - choice of statistical techniques - communication of design, analyses and results.
Independent researchers and their doctoral students can fully outsource data analysis and get advice on how to deal with reviewers and how to write the statistical analysis plan in research proposals.
Research groups could request for automating data collection and processing as well as tutoring (workshops, group discussions, ...).
Consultations are free of charge but the number of consultations can be restricted by the consultant when the client is unwilling or unable to put in the necessary efforts.
The consultants will offer advice to the best of their abilities. If the requests are beyond the expertise of the consultant, options are discussed and clients may be referred to elsewhere. No responsibility can be assumed for any inadequate advice given if based on false or incomplete information from the client.
Suggestions from the client as well as attempts to perform the analysis are discussed in terms of advantages and disadvantages in a way that would meet the background knowledge of the client.
A single consultation to offer advice or discuss an analysis takes up a maximum of one hour. Consultations take place at the consultant's office and can exceptionally be done with skype calls if otherwise impossible and when motivated sufficiently. A consent of the promoter is in principle required for (Phd.) students.
Data Analysis can be free of charge, but a consultant can accept or refuse the analysis of data dependent on the available time in relation to the perceived importance of the study. Researchers are advised to request for assistance well in time, no guarantees are given on immediate support. For a more involved commitment, for example large scale projects, researchers are advised to discuss the expected contribution and suitable compensation well in advance. Such involved commitment would ideally be part of a project proposal to attract the required funding.
Co-authorship includes a commitment to help with the communication of methods, statistics and results, as well as to the review process.
Interfaculty Center Data processing & Statistics, Vrije Universiteit Brussel
The quality of the data in relation to the research questions of interest defines the statistical potential and are the responsibility of the client. A consultant only analyses data to address clear hypotheses offered by the requesting researcher. A consultant does not read through substantive literature, this remains the responsibility of the requesting researcher. No responsibility can be assumed for any inadequate analysis if caused by false or incomplete information from the client. Consultants can decline or stop any assignment based on poor quality of data, research questions that are considered inadequate, or the refusal of clients to take methodological concerns raised by the consultant into account. Data offered to ICDS should be in an appropriate format. ICDS should be contacted in advance for assistance if in doubt. Data remains property of the researchers or any other person who delivers the data. Data will be saved in an internal ICDS database. Public use will require consent of an owner of the data.
While an attempt is made in advance to agree on an appropriate timing for performing the data analysis, the time required typically remains uncertain because of various potential complicating factors, for example a change in statistical technique following observed data properties. A consultant can not be pressured into a quick delivery of the results (time) nor into the use of specific techniques (action).
The consultant allows for sufficient time to interact with the client to cover the assignment in full detail, including all issues related to the design of a study, the data and any foreseeable methodological challenges. The consultant informs the client about possible conflicts of interest, ethical limitations and all other aspects that could impact the design, method and analysis. In case of missing data, extreme outliers, or other unexpected observations, they should also be discussed so that the consultant can deal with those in an optimal way.
The client will in the end obtain all relevant details on the pre-processing, coding of variables and used software. In preparation of scientific communication the consultant will point out conclusions but also limitations and conditions for those conclusions to be valid. The consultant will at the request of the client read through the relevant sections and where necessary suggest improvements.
Tutoring and Automation should be discussed well in time and is accepted at the discretion of the tutor/analyst/programmer. Because of the commitment involved, an appropriate compensation should be discussed in advance. The objectives should be specified in detail by the client and a course of action agreed upon.
Co-authorship and Acknowledgments
The client can evaluate the contribution made by the consultant, and appreciate it explicitly in resulting scientific output. A substantial contribution, typically involving a data analysis, in general would result in offering co-authorship. A minimal contribution, typically involving advice, could result in an acknowledgment.