My experience is different from previous comments. My business is to help undergraduate and graduate students in good, correct usage of statistics in their research. It is definitely a difference in what science discipline is student doing his/her research. In many cases, I directly did a statistical design research and all statistical calculations and the interpretation of results. Of course, it is always necessary to instruct them in the calculation procedure. I believe that the high degree of specialization in science brings also the need for utilize the services professional statistics.

Hi Debra,

Is it acceptable for one to delegate a task ? For a professor to delegate coding tasks to an assistant  in the research process ? From my modest student, but soon professional, position, I would say it is, except if you are a student. Studies are a period of life, I believe, where you are not trying to be as efficient as possible (or to make as much money as possible) but on the contrary, you try to learn as many things as possible to prepare yourself for the rest of your life, be it a business or an academic life.

Rather delegating, I would consider taking the time to face the difficulty, in this case the issue of coding and interpreting the test, in order to understand it better in the future.

Best regards,

Geoffrey

The student must understand and be able to manage doing it by him/herself.  This is especially critical if s/he desires to remain in academia.  To make the student understand and appreciate the stats, we just have to go through the processes using the formulas and using a scientific calculator, just one that has a square root function.

I agree with the above answers; the purpose of a research graduate degree is for students to get training in all aspects relevant to doing research in their field. If statistics is part of that, then the student should be learning how.

Hi Sharon,

In my opinion, delegating the task totally to a professional in the absence of the student is highly unethical. It is quite different from allowing a seasoned professional handle your stats and explained explicitly in your presence. But the best practices is to learn how to it yourself. You wont need to consult a professional again if you want to do a similar thing tomorrow. 

Let me share an experience. While an undergraduate student doing my project then, I was faced with a challenge of performing a statistical analysis of my work. I downloaded some free statistical software trial versions from internet like STATISTICA and co but couldnt use it after spending endless time, data and energy to do so. It wasn't fruitless because it made me to understand the importance of a professional. Well, because of my curiosity to do it myself, A fellow (postgraduate student then) I discussed my problem to surprisingly introduced me to SPSS software. He thought me how to install it, run my analysis while my supervisor thought me how to interpret the result from the analysis. After overcoming the task, i became a "statistician" immediately because some students came from even other departments to seek for my help. I am glad I learnt and it was one of the best experiences while further equipping me for more research.

From the shared experience, no reason and on no account shall students delegate their stats analysis or contract it totally in their absence. If they do so, they may not have the opportunity of learning it again. When such a graduate finds its way as a Teacher/lecturer into educational institutions, I wonder how he will supervise his students in this regards.

Best regards,

Egbuna, C.

My experience is different from previous comments. My business is to help undergraduate and graduate students in good, correct usage of statistics in their research. It is definitely a difference in what science discipline is student doing his/her research. In many cases, I directly did a statistical design research and all statistical calculations and the interpretation of results. Of course, it is always necessary to instruct them in the calculation procedure. I believe that the high degree of specialization in science brings also the need for utilize the services professional statistics.

University administration should check and stop such activities

As a teacher, I am not OK with this. The best way to deal with this situation is to train such students by doing several stat problems and then to examine them.

Dear Debra:

In my view, a professional who has to use statistical tests needs to know the principles and criteria on which those techniques are based.

Some tests are really complex and in these cases is appropriate to have the advice or help of a statistician.

However, the only way for students to understand the statistical principles is to run the tests by themselves. This does not mean that they don't have the help of an expert if necessary. 

I believe that it is acceptable in some case as long as students understand and can explain the approach. Students need to be encouraged to learn but they definitely need help. Nobody knows everything. That's why there is a biologist and a biostatistician.

That is wrong,during graduate level, the students should be exposed to hands-on experience with Statistics may be a smaller level, so that they can understand and appreciate the application of statistics and its benefit.

For me even use of software packages such as SPSS is not advisable for graduate level. However, when more data and bigger projects are part of Post-Graduate and higher education, then Software support is required. 

In both the cases the student has to do by themselves. Facilitation by others will handicap them in the future  (vulnerable to depend on others for support) 

Krishnan, why vulnerable to depend on others for support? Good research requires collective cooperation - in graduate level too. What it is a good cooperation, if not included in the research team also an expert on statistics?

It should be acceptable, if the student know the applicability of the technique and relevant applications.

Yes, but the statistician should be a good expert or doctor.

Mariano, you´re absolutely right. Even among experts on statistics are specialists in economics, macroeconomics, microeconomics, engineering, construction, etc. etc. And then there are specialists in clinical epidemiology, laboratory medicine, pharmacoepidemiology, etc. etc. For example, I myself a few years ago asked for help colleague statistician, an expert on statistical methods in meta-analysis.

Certainly graduate students should be aware of the challenges, pitfalls, and advantages of using particular methods of statistics for analyzing their data,  However, they do not need to become experts.  If the research and the ultimate innovation is primarily laboratory based, then the use of proper statistics is simply a tool, and the student should be taught to wield it, but not necessarily to be able to invent it.

I disagree with those who expect the student to be embargoed from technical assistance.

Consider the statistician analyzing data from the literature or from a colleague's laboratory.  Do we expect him or her to be able to construct or operate a mass spectrometer?

The student should be allowed/expected to be creative and innovative in some important phase of a broader endeavor, but should be encouraged and assisted in ancillary tasks - this is why we have committee members and faculty.

In my opinion, it is absolutely understandable a weak graduate student, with regard to the running of a chosen statistical test, to ask for a help on it.

But, after the end of this running, he must be able to run a similar statistical test alone. Only in this case, the help would be really useful to him in the future.

It is good and ethical to help a prospective colleague in his start, especially when there is a real need.

I think  it is acceptable unless it is in  the statistics itself 

Dear Debra,

Thanks for the invitation. I utterly concur with Dr. José Eduardo Jorge. In my opinion, graduate students can have the help of a relevant expert in their field, but they shouldn't have someone else to run their statistical tests for their study data.

Best regards,

Mehdi

Is it acceptable for graduate students to have someone else do their stats?

There are incidents that some graduate students "outsourced" their statistical analyses to someone else.  Even though they might not know the detailed technicality but they know what the statistical tests to do what etc.  In my opinion, the graduate students should perform the statistical analyses by themselves - reason being:

Dear Debra Sharon Ferdinand,

Very good question. Thank you for stimulate this debate 

Nowadays it's easy to learn anything

You can find Help or good advice on Line And turorials .., with YouTube for example or moocs

So student can improve And learn by himself

Best 

Worldwide, the tens of thousands universities is active. The exact number is 23 729, which were included in the Ranking web of Universities for the year 2015. Certainly, the level and quality of graduate education and graduate students are different. That is why I very much welcome an open discussion on the topic launched by Debra Sharon Ferdinand. There were given different views and different experiences from different universities. Well thank you!