How can I use open data based on interviews?

Question

Qualitative data and interviews in particular have the issue of containing sensitive material. Ethical issues have historically prevented people from releasing interview transcripts (it is difficult to question this). Even if the participants agree to have their data go open, it is an expensive activity to go through the transcripts to anonymize the data. Given that ethical issues might still arise, one does not feel motivated to share qualitative data.

I myself have recently conducted a qualitative study based on interviews, in press at an open access venue, but had to regretfully comment that I could not release the data because of its sensitivity (and the participants were not confident to share their related data).

So, my question is: how can I do open science in interview-based studies, where the data is qualitative?

I think that this question differs from the recently posted How to deal with sensitive individual data in open science?, as it is more about opening up interviews and qualitative data in general than on privacy of the data.

This post has been migrated from the Open Science private beta at StackExchange (A51.SE)

Comments

If you think that this thread should be migrated to OpenData or another SE site because the OpenScience beta is closing, please edit the list of questions shortlisted for the migration [here](http://meta.openscience.stackexchange.com/questions/73/).

This post has been migrated from the Open Science private beta at StackExchange (A51.SE)

Answer 1

I think that the easiest way to deal with the interview problem you mention is to not collect any personally-identifying data in the first place. It is far easier to initially design for releasability and only collect non-identifying information than to go back afterwards to remove it, as you pointed out.

For example: you could assign each participant a number for identification purposes (that you could provide to them in case they need to contact you about their interview) and not collect names, addresses, phone numbers, etc. Obviously, there are cases where this would complicate a study but, in most cases, you should be able to gather the data that you need for your study without collecting anything sensitive.

If you must collect sensitive information, you should keep it in a separate location that will not be released. In that location you could correlate the participant identification number with the necessary sensitive data.

This post has been migrated from the Open Science private beta at StackExchange (A51.SE)

Answer 2

For the science to be open, it needs to have the methodology published along with the results. This means that the survey itself has to be properly documented, in order to be reproducible.

This kind of work is often done within the social sciences community - particularly the International Household Survey Network (IHSN). They use a metadata schema of the Data Documentation Alliance (DDI), which has documentation tools.

These documentation tools may help you to properly document the questionnaire or interview - this will help reproducibility (ignoring the privacy issues for now as you suggest).

Perhaps starting the study with tools which allow reproducibility or some form of verification instead of post-actions (which you say can be expensive) might reduce the overhead. Although not an expert in this field, I think the DDI suite might help with this.

what is more, publishing the survey along with the results is also good practice. Nesstar publisher and the Nada catalogue publisher can do that.

This post has been migrated from the Open Science private beta at StackExchange (A51.SE)

Comments

Following typical StackExchange style, it's probably better to include some of the details from the link here rather than point people elsewhere.

This post has been migrated from the Open Science private beta at StackExchange (A51.SE)