GRS ‘StatsHelp’ Program
Don't leave it until the last minute to involve a statistician in your Project!
Research theses and research projects sometimes fail or require substantial revision due to:
- Poor experimental design
- Incorrect application of statistical tests
- Violation of assumptions, and
- Incorrect interpretation of results.
Data collection, analysis and interpretation of statistical output can be an expensive exercise in both time and money. It is tragic if a research student’s thesis is failed because of such flaws, which can be irredeemable. Involving a statistician, preferably before you start collecting data, can minimise the risk of any of these problems occurring.
Take advantage of the StatsHelp Program, funded by the Graduate School of Research.
Any JCU student who is enrolled in a:
- PhD
- Research Masters
- Professional Doctorate, or
- Honours and coursework Masters Degree (fee basis only and pending availability).
Supervisors are very welcome to attend consultation sessions with their postgraduate research students.
The StatsHelp Program is funded by the Graduate Research School. Access to the StatsHelp Program by enrolled postgraduate research students is free.
In the event that a student requires urgent help with a statistical problem, an appointment can be scheduled with a statistician usually within several working days.
StatsHelp appointments, or consultations, may be booked for up to one hour.
The first half of the consultation usually requires that the student explains their problem or project objectives to the statistician. The statistician may then ask questions to clarify they understand the problem to be solved. The second half of the consultation involves the statistician explaining the different statistical approaches that could be applied to the situation and the student asking questions about these different statistical approaches to ensure they understand the concept.
To make the most of your appointments with the Statistician, be prepared. Students who come to the meeting poorly prepared often spend too much time explaining their situation to the statistician, thus leaving less time for the student to learn more about the statistical techniques they could apply to their data.
Where possible, research students and, if desired, their advisors should register to attend one of several statistics training workshops offered through the Research Skills Program each year. For further information and training dates, visit the Professional Development Program webpage.
One hour may seem like a long time but it passes very quickly. Students who are well organised and prepared for their consultation will gain the most from their session.
What is/are your research question(s)?
Write down your research questions and hypotheses. Note, StatsHelp statisticians can also provide advice regarding the design of questionnaires.
What data do you have?
Summarise what data you have or plan to collect that will be used to answer your research questions/hypotheses. Note the variable names, type of data, number of samples, replication, etc. Diagrams/pictures that describe how/where you will collect your data can also be helpful.
Graph your data
If you have already collected your data, it is highly recommended that you graph the data and bring along a hard copy of these graphs to the meeting.
Effective communication
It is important to remember that the consulting statistician is not an expert in your field. You will need to concisely explain the concepts associated with your data and data collection with minimal use of jargon.
Search the literature relevant to your discipline
Different disciplines and journals will favour different statistical approaches. It is good practice to search the literature to understand how other researchers have applied a statistical method to a situation that is similar to your investigation or target journal.
The statistical advice given will be based on your level of understanding. You will be advised of a statistically correct technique, but not necessarily the optimal statistical technique. You will be advised in writing if a compromise has been made.
The statistical consulting team is able to answer most queries presented by students. From time to time, however, some cases require specialist input that exceeds the role of the consulting team. In these situations, the consulting team will do its best to refer the student to relevant information sources.
As mentioned above, statistical short courses are offered annually as part of the Graduate Research School’s Professional Development Program.
Registrations for Research Skills courses fill up quickly, so be sure to register early. Students who register to attend Research Skills courses but don’t show up on the day (a ‘no-show’) or cancel at the last minute without a valid reason will waive their access to the free one-on-one statistical consulting sessions.
If you require more advanced or time-consuming assistance, the statisticians may be willing to enter a collaborative arrangement which will result in co-authorship of any paper which uses our advice.
As part of your initial enquiry, please provide:
A brief background of your research project;
Details of your chosen research question; and
Details of your chosen statistics software.
Please also provide advice as to the problems you are experiencing with your data, or the questions you would like to ask the statisticians.
If you are seeking assistance following the review of a submitted manuscript by external reviewers, please provide a copy of that manuscript along with the reviewers’ comments.
Send an initial request for a consultation, along with the details outlined above, to GRS@jcu.edu.au.
Availability
The consulting statistician(s) will specify dates and times they are available for consultation. Dates and times are usually fixed. StatsHelp team members are generally available throughout the working week on the Townsville campus, subject to personal workloads. Assistance is provided for students based at the Cairns campus by only one StatsHelp team member, who is available on Thursday and Friday each week (you must make an appointment).
Book in advance
It is not uncommon for the statistical consulting service to be booked out several weeks in advance. If you would like to meet with a statistician prior to an important date, e.g. your Confirmation of Candidature seminar or data collection, remember to book several weeks in advance.
The StatsHelp program is offered in support of JCU’s graduate research students. Students who book an appointment and then do not turn up at the agreed time without explanation are likely to forfeit any further assistance from the StatsHelp program. Help us to help you – arrive on time, be prepared, and make the most of your StatsHelp appointment.