I have been involved in projects that have used assessment data from students for research. In one case, a New Zealand university gave us approval to use any naturally occurring data while another refused outright saying we needed volunteers. It seems to come down to individual ethics committees. Here, in the UK, I am given the impression that I should be able to use naturally occurring (i.e. assessment items) and I have with the assistance of a student gained approval to run research assessments in a tutorial session.

The key issues tend to be:

1) maintaining anonymity of the students in the research process. This can be achieved by ensuring any identification of the student is removed from material passed into the research process. Preferably have someone other than yourself carry out this process.

2) Gaining approval from students for the use of their data in research (informed consent). In the case of using naturally occurring data, the students were advised at the beginning of the course or module that any assessment items may be used in research. They were given access to information on what data would be used and how it would be used. They were given the opportunity to withdraw but none did.

3) Allowing students to volunteer or ask to be removed from the research. This is a little more difficult and is the reason why the university above declined. However, it is easy enough to set up someone to act as the agent who managed the list of those who have agreed to participate and to have them remove assessment items that cannot be used in the research.

4) All of these processes are about ensuring that no harm or disadvantage occurs as a result of participating or not participating in the research. However, my argument is that we should be using naturally occurring data to validate pedagogical practice and we may be causing more harm because we fail to do such research and analysis. So much teaching is done based on opinion rather than based on evidence.

5) Once approval is achieved, the next issue is to ensure that assessment items actually provide meaningful data for research. Here I would argue most assessments do provide data that will help us understand pedagogical issues however, it may be helpful to design specific question that can be analysed using a specific research method. This is what we did with research into reading, tracing, and writing program code. The research was quite successful.

I hope this helps or at least helps you think through the issues.

You are right about pedagogical issues not counting as research within the disciplines but educational researchers are always looking for ways to become involved in discipline specific research. If you have an education department talk with them and see whether you can find a researcher who might be willing to work with you. This might make it easier to gain approval and also give you are wider publication base that simply disciplinary based education publications.

The more info you collect, the more likely you are not to overlook a variable that has influence. To balance against this, there must be a glimmer of expectation that the variable will have a measurable effect. Why record if the student is right or left handed, when there is no reason to expect an effect?

Jayvee,

Well, this question has more than one side to it:

(1) if we are talking about research involving collecting data from or about students, at our university - this is considered a very high risk research - and we can only commence collecting data if we have the approval of the highest ethics officer at the university level.

(2) the research on pedagogical issues is great - however, with the recent concentration on publications in the discipline - the academics are encouraged to concentrate their work on the discipline rather than pedagogy or curriculum - as 80% of publications in the year need to be in the discipline...

(3) research involving students - is not looked at highly when it comes to publishing.

Regards

Theodora Issa

I have been involved in projects that have used assessment data from students for research. In one case, a New Zealand university gave us approval to use any naturally occurring data while another refused outright saying we needed volunteers. It seems to come down to individual ethics committees. Here, in the UK, I am given the impression that I should be able to use naturally occurring (i.e. assessment items) and I have with the assistance of a student gained approval to run research assessments in a tutorial session.

The key issues tend to be:

1) maintaining anonymity of the students in the research process. This can be achieved by ensuring any identification of the student is removed from material passed into the research process. Preferably have someone other than yourself carry out this process.

2) Gaining approval from students for the use of their data in research (informed consent). In the case of using naturally occurring data, the students were advised at the beginning of the course or module that any assessment items may be used in research. They were given access to information on what data would be used and how it would be used. They were given the opportunity to withdraw but none did.

3) Allowing students to volunteer or ask to be removed from the research. This is a little more difficult and is the reason why the university above declined. However, it is easy enough to set up someone to act as the agent who managed the list of those who have agreed to participate and to have them remove assessment items that cannot be used in the research.

4) All of these processes are about ensuring that no harm or disadvantage occurs as a result of participating or not participating in the research. However, my argument is that we should be using naturally occurring data to validate pedagogical practice and we may be causing more harm because we fail to do such research and analysis. So much teaching is done based on opinion rather than based on evidence.

5) Once approval is achieved, the next issue is to ensure that assessment items actually provide meaningful data for research. Here I would argue most assessments do provide data that will help us understand pedagogical issues however, it may be helpful to design specific question that can be analysed using a specific research method. This is what we did with research into reading, tracing, and writing program code. The research was quite successful.

I hope this helps or at least helps you think through the issues.

You are right about pedagogical issues not counting as research within the disciplines but educational researchers are always looking for ways to become involved in discipline specific research. If you have an education department talk with them and see whether you can find a researcher who might be willing to work with you. This might make it easier to gain approval and also give you are wider publication base that simply disciplinary based education publications.

The cost of gathering "naturally occurring" student data is minimal.

Jayvee, it's not clear whether you're asking about collecting data **about** graduate students, or **about** undergraduates, that would then be used by professors who teach in a graduate program (but who would be using data **about undergraduates**, that was collected by the professors' graduate students, for instance, teaching separate lab sections of a course.) That is, are you trying to conduct **graduate-level pedagogical research,** or **undergraduate-level pedagogical research** with the assistance of graduate students? Could you please clarify? It makes a difference in the relevant responses to your question.

For improving practice. It is beneficial to base pedagogical research on students data, so we can reach what we call best practices or research based practices.

When you work in education, as I do, it is very convenient to carry out research on the student population with whom you come into contact and also, I feel, informative in terms of improving university pedagogy. Here in France we do not, as yet, have to comply with ethics committee conditions in order to carry out research into these matters. However, I do make sure that I inform students about the research I am undertaking, guarantee their anonymity, reassure them that the research will have no impact on their marks and allow them to opt out.

Doing educational research on your students is most convenient. You do not have to travel far to get your data, and you can repeat the experiment easily if necessary. The normal rules of ethics apply. You get no staff discount on that!

Errol I found your comments very beneficial, thank you!

We certainly do research on students at our universities:

http://reflectivewriting.concordia.ca/

Usually it is better to do research on a colleagues class rather than your own.

If you are doing research on your own class it is essential that interviewing and other such interventions not be conducted by yourself and that students are assured that you would not look at the data until after the exams are submitted. Other ethical considerations have already been mentionned.

Jayvee,

In our institution we have some interdisciplinary researches involving students.

All depends on the type of research you want to do.

First, remember that all research must be submitted to a institutional ethics committee and student participation should be voluntary and anonymity provided by the researcher.

There are several issues that can be addressed as assessment of performance in relation to specific teaching, assessment of knowledge on varied topics, among others.

If applicable the educator must use all possible variables on a particular group always taking statistical careful not to create overestimated results.

I think it depends on what you are researching. For example if you are trying to explore students experience of learning eg level of satisfaction, Despite students responses being anonymous , it would be hard to get honest and open feedback due to the relationship with the researcher (educator). I think in this case, educators should not use their own students (even if they get ethical approval). It is well documented in the pedagogical literature how the relationship of the researcher can bias the results (usually, favourably for students fear of causing offence) but I'm not sure of what you would be researching.

On the other hand if you are worried about supervisors using your research data for their own research, that's a different matter ?

We have conducted action research involving our students.We wanted to find out if a certain method of teaching have an effect to their performance and in what way can we improve that certain teaching method.

As I work in a centre for university teaching, our students are the main focus of our research. Eveything has to be checked very carefully to ensure we follow ethical guidelines, but that's how we improve our performance.