Client’s guide to the Statistical Consulting Lab

We are prepared to render assistance at every stage of the research process, from planning a study, through collecting data and statistical analysis, to interpretation of results. Whether a particular service is appropriate depends upon the context of the consultation. Our central focus is on statistical methodology.

Any faculty, staff or graduate student working on research (not homework) is welcome to contact us. There is no charge for our service. However, as the Statistical Consulting Lab (SCL) is funded in part by the School of Forestry, first priority is given to faculty, staff and graduate students in Forestry. Generally we are able to accommodate all requests for appointments from members of the University community.

We expect our clients to have a working knowledge of basic statistical concepts and methods. What you need to know depends partly on the nature of the research that you are undertaking. Graduate students who will be engaged in research are strongly recommended to take STA 570 (and preferably STA 571 as well).

The general answer is yes, but there are qualifications. The significant word here is ‘statistical’. SCL will provide assistance with the use of statistical computer software, but we do not provide assistance with more general computer software, such as operating systems, word-processing programs, and spreadsheets.

No.

It is appropriate for a consultant to discuss with a student alternative experimental plans (as well as analyses) that may be suitable for his or her research; to help a student correct a computer-program setup; or to help a student to interpret the output from a statistical computer program. It is the policy of the SCL that the student’s thesis/dissertation adviser be notified when a student uses the SCL.

Although there is a large grey area demarcating activities that are clearly appropriate from those that are clearly inappropriate, it is generally not proper for a consultant to: make decisions about how a student should analyze his/her data; analyze the student’s data; or write computer program setups for the student. We realize that at times a consultant may need to examine a student’s data, and that students may require relatively more assistance with forms of statistical analysis that are unusual or difficult. We have a wider latitude in assisting faculty and staff with their research. Due to limitations in our resources, we are unable to provide data entry and data management services. However we will provide advice on how a data file should be structured for most common computer-assisted statistical analyses. The extent to which SCL staff work with faculty on a project will in part depend upon whether the relationship is collaborative or consultative. Implicit in this distinction is the notion that a collaborative relationship will generally mean that the SCL staff member serves as part of the research team, plays an active role in the research study (particularly in the design and analysis phase, and perhaps in the problem formulation stage as well), and in the joint publication of the research.

Email us at statlab@nau.edu for an appointment. We are located in Room 116 in the Adel Mathematics Building.

Please email a copy of our contact information form before your scheduled appointment time.

From the beginning of the project! A well-designed project can save time, money and maximize the information obtained. We have seen cases of poorly-designed studies from which NOTHING was salvageable. Collaboration between the consultant and the client will be far more fruitful if the consultant is involved from the beginning of the project.

Advice concerning statistical design and analysis is not available over the phone. In our experience a ‘quick’ question often masks a complex research problem, which requires a consultation for discussion of all the issues. However, we do provide consulting services via video conferencing for those clients who are unable to travel to the Flagstaff campus.

After you and the consultant introduce yourselves, the consultant will ask for information about your research and ask you what you would like to accomplish during your first session. It is helpful for you to prepare a summary of your research, and in some cases you might consider providing this information to the consultant in written form.

Often it helps to summarize the design of a study in a diagram or a table. It is generally useful for the consultant to understand the purpose and context of your research, rather than viewing your work as an abstract statistical problem. An initial consultation usually takes from one-half to one full hour.

You are welcome to return as frequently (within reason, of course) as your research requires. Many clients have their questions answered completely in a single consultation, but often more contact is required. Frequently, a consultant must do work outside of the consulting sessions to answer a client’s questions.

You may acknowledge our assistance in the usual manner if and when your research is published. We would appreciate receiving copies of research that has benefited from our assistance because documentation of the fruits of our labor is useful to us in renewing our funding.

In instances in which a consultant’s contribution has been central to a project (e.g., in which a long term collaborative relationship has been agreed to and undertaken), co-authorship on the resulting publication is a reasonable expectation.