University of Tasmania, Australia

Human Research Ethics (HRE), evolved in response to some of the appalling medical 'experiments' of the 20th century and was designed to protect participants involved in medical research. It has since expanded from its original narrow focus on medical research to cover social science research – and is now expanding still further to include research fields such as the visual arts, music and journalism.

You can find an excellent summary of the evolution of HRE in Australia on the NHMRC's website at: http://www.nhmrc.gov.au/health-ethics/human-research-ethics/history-ethics-and-ethical-review-human-research-australia.

HRE is guided in Australia by the 'National Statement on Ethical Conduct in Human Research' 2007 – and a very easy-to-use html version of this document can be found at: http://www.nhmrc.gov.au/book/national-statement-ethical-conduct-human-research.

The National Statement is the required set of guidelines for all research ethics applications. Its recommendations guide the activities of all HRECs in Australia – whether they be based at a university, a hospital, a government agency, a local council, or a not-for-profit organisation – the HRE rules you are asked to follow at UTAS are based on these national guidelines.

Research Integrity, by contrast, is what many people think of as 'ethical researching', and this name was adopted because the term 'research ethics' was already being used by the National Statement (and the equivalent animal research ethics codes).

The term 'research integrity' in Australia refers to a set of rules (codified in the NHMRC's 'Australian Code for Responsible Conduct in Research' – and available from: http://www.nhmrc.gov.au/guidelines/publications/r39), which cover a variety of research activities: management of research data, supervision of research trainees, publication/dissemination of research findings, authorship, peer review, conflict of interest & collaborative research across institutions.

It is worth nothing that some of the research integrity requirements (most notably management of research data, peer review and conflict of interest) are incorporated into the HRE application forms and need to be addressed by those seeking HRE approval.

Section 7 (Funding): having ticked 'not funded' you still need to explain who will take responsibility for any expenses which might arise. Of course, there might not be any expenses, but the Committee is still required to ask for a 'responsible person/group'

Recruitment of Participants (Minimal Risk – Section 8; Full Risk – Section 10): 'classic' snowball sampling, where you ask the people you do know to give you the contact details of other people who might be suitable so you can phone/email them directly, is now a breach of personal privacy under the Tasmanian PIP Act (2004). But don't despair – there is a (slightly longer) alternative which will still allow you to set a snowball rolling. You simply confirm in this section that participants will not be asked to provide you with the personal contact details of other potential participants, but will instead pass along information about the study to other people who can then contact you.

Data Identifiability (Minimal Risk – Section 9; Full Risk – Section 11): it's possible the way this section is described has confused you – we are seeking to learn what sort of raw data you will be collecting, rather than asking you about the form your data will take once transformed by your analysis processes. If you know who your participants are (e.g. you are emailing them to arrange a workshop or focus group, or asking them questions which identify them) your data cannot be non-identifiable but must rather be, at least, re-identifiable.

This is not necessarily a problem – we simply need you to explain how you will be protecting the identities of your participants (perhaps by using a table in which real names will be matched with pseudonyms which will thereafter be the only form in which identities are referenced?). If you are taking the 'table' approach, you will need to explain that only the researchers will ever see the linked names/pseudonyms and that these will be stored separately from the rest of your data.

If your participants are identifiable (i.e. if people reading the thesis or article in which you publish your findings can identify your participants, despite your attempts to hide their identities behind pseudonyms) then you must also offer these participants the opportunity to review their transcripts and, ideally, anything in the thesis/articles which might identify them. Participants who consent willingly to be identified, provided they can check what's said about them before it's published, are giving truly informed consent.

An extremely important consideration in the ethical conduct of a human research project is the degree to which the voluntary and informed consent of participants is obtained. The design of a project must provide an appropriate mechanism to provide such consent, and an application for ethical clearance for a project must identify and address any potential issues that might have an impact upon the voluntary and informed nature of the consent obtained from participants.

In the case of students, what the National Statement calls their 'unequal relationship' with teaching staff (see Ch. 4.3 of the NS) means that three types of issues can directly affect the nature of their consent:

1. Perceived impact upon grades or service – Where a student perceives their decision about participation might affect either the grades they receive or their ability to access services, this may result in significant coercive pressure.  Such considerations can be especially acute where the researcher is also their tutor or lecturer (where the data collection is not anonymous, so their participatory status will be known to the researcher); and / or where the data reflects upon their learning or attitudes towards the unit being taught;
2. Perceived impact upon relationship with tutor or lecturer – Where a student perceives that their decision about participation might affect the way their tutor or lecturer perceives them, this may result in significant coercive pressure. Such considerations can be especially acute where the students' participatory status will be known.
3. Peer pressure – "Open" recruitment or data collection which occurs in front of an entire cohort of potential participants (e.g. a tutorial session) can result in peer pressure that may have a coercive impact upon potential participants.

The recruitment and informed consent materials must stress the voluntary nature of participation and explain very clearly indeed that participation will not have any impact on grades or a student's ability to access services. If the researcher is also a tutor or lecturer of the potential participants, it would be preferable that they remain unaware of who participates.  Where this is not practicable, the researcher must go to great lengths to assure potential participants that they are under no pressure to participate. 

It is generally preferable not to conduct recruitment or data collection in a situation where an entire cohort of potential participants are "put on the spot" and asked to participate in a research project.  Potential participants should be given time and space to consider their participation, especially where there are potential coercive pressures upon their decision.

If your survey is intended to be anonymous it should ideally remain anonymous for all respondents, even those who choose to participate in later stages of this study (otherwise you will find it very difficult to treat their survey responses in exactly the same way you treat the other, anonymous responses – human nature being what it is).

The easiest way to keep your survey anonymous but still invite participants to identify themselves for later research tasks is to remove the link between names and questionnaires for those respondents who are willing to participate in Phase II. Rather than including the space for contact details on the questionnaire itself, you can simply provide a link to a second survey questionnaire where respondents can include their details. This is very easy to do with all the major online survey providers – and can be handled by including a separate, detachable sheet with your hard-copy questionnaire, if you're (also) using paper-based documents. This will still give you all the details you need, without the requirement to link the two survey forms together, thus ensuring anonymity for all respondents

I'd recommend avoiding the use of userids completely, because even asking participants to make use of their userid to access the survey environment means that you will know who has considered taking the survey. Since it is unlikely you will obtain a 100% response in any case (I don't think I've ever come across a 100% response survey!), I would recommend that you accept you won't know who has completed the questionnaire and, instead, simply ask participants to go to a particular URL and take the survey – this is the approach used by most researchers. It's especially important if you're planning to ask 'tricky' and/or very personal questions!

Quite apart from the ethical benefits this approach provides, you will find that more respondents are willing to participate in your survey if they can be assured of genuine anonymity …

I can quite see why you want to be able to link participants with outcomes so that you can test hypotheses – and I also understand why the student id is the most readily available means of obtaining such a linkage. Nonetheless, using student ids in this way means that your data will be identifiable – rather than re-identifiable – and this will lead you into much more onerous requirements in terms of consent, e.g. you will need to explain that participant responses will be identifiable in your PICFs which will require tick-boxes for approval, etc.

Unless you really need such a linkage, a simpler solution is to take the 'pseudonym' approach to protecting the identities of your participants – perhaps by using a table in which real names will be matched with pseudonyms which will thereafter be the only form in which identities are referenced? If you do decide to make use of such a 'table' approach, you will need to explain that only the researchers will ever see the linked names/pseudonyms and that these will be stored separately from the rest of your data – but this will still be considerably easier for you than the requirements of your present approach …

Section 11 of the minimal risk form asks me 'where is this project to be conducted' – what's that all about?

This question relates to researcher safety. The University is not only responsible for the welfare of research participants, but for the welfare of the researchers themselves – and when researchers plan to gather data in locations where they may be at risk, the University is held responsible by its insurers for any accidents which might occur. Researchers who plan to gather data in such locations are therefore asked to provide evidence that they (and their supervisors, in the case of a student project) have developed a suitable safety protocol, e.g. that the researcher will carry his/her mobile phone at all times, will notify a supervisor or colleague (or some other suitable person) when data gathering is about to commence and will call or text when the data gathering visit is safely completed

We ask for such a protocol in the case where a researcher plans to gather data in participants' homes, as well. While this may seem over-zealous on the part of the HREC, 80% of accidents occur in homes and, while the University is responsible for researchers' safety, it is better to be overly cautious than not cautious enough …

Monitoring (Minimal Risk – Section 12; Full Risk – Section 19): this relates to what the National Statement describes as "the process of verifying that the conduct of research conforms to the approved proposal" and is the responsibility of the institution which approves the project (in your case, the SSHREC). You need to explain both:

• how the project team itself will be monitoring the project; and
• how you will handle reporting to the SSHREC.

Essentially, we are asking you to confirm that you will provide annual reports; and will immediately notify the HREC of any adverse incidents.

We are asking you to confirm that you will provide annual reports; and will immediately notify the HREC of any adverse incidents. But if you're having difficulty working out how to express this in your application form, here are some good examples which have been slightly amended from previous applications:

Example 1: The chief investigator, <name>, will monitor the conduct and progress of the project.  The researchers will conform to HREC reporting requirements as indicated in NS 5.5; will prepare a final report; and will immediately notify the HREC of any adverse events.

Example 2: This project will be monitored in accordance with Section 5.5 of the National Statement. Annual and final reports will be submitted to HREC Tasmania when due, as will reports of any adverse events associated with this study. Regular meetings will be held between the student investigator, the chief investigator and the co-investigator.

Example 3: This research project will be conducted in compliance with the approved proposal and any conditions attached to this. Unexpected occurrences, events, outcomes or disruptions that may arise during the research will be reported immediately by the CI to the Integrity and Ethics division of the University of Tasmania. On completion of the research <date> a final report will be submitted to the Integrity and Ethics division of the University of Tasmania. 

Data Storage (Minimal Risk – Section 13; Full Risk – Section 21): most people seem to feel comfortable explaining where they plan to store their hard copy data, but often forget to clarify how electronic data will be stored. Essentially, you need to explain that these data will be stored on password-protected UTAS computers (or, more probably, on your School's network drive)
I'm going to be gathering my research data [overseas / on the mainland/ in regional Tasmania] and I can't work out how to store my research data safely while I'm on the road

Data Storage (Minimal Risk – Section 13; Full Risk – Section 21): both the Australian Code for the Responsible Conduct of Research, as well as the brand-new UTAS Research Data Management Policy and procedures, require that electronic research data be held in secure, regularly backed-up storage within the School in which the candidate is enrolled. In view of the advent of UTAS cloud storage, however, this rather narrow definition has been extended to include the MySite facilities (see http://www.utas.edu.au/service-desk/mysite which explains how to link to a student cloud storage site from a UTAS laptop or desktop PC; or http://www.utas.edu.au/it/application-and-service-hosting/collaborative-file-storage-students which explains how to access student cloud storage from an offsite location). A successful application will generally include one of these more secure and reliable PhD student storage facilities for your electronic data
I'm finding it hard to work out just what I can and can't do with my research data – can you help, please?

Here is a quick list of do's and don'ts for research data storage:

• Physical data must be stored in a secure, locked facility within your organisational unit – not at home, though you can store a working copy of your data at home
• Electronic data must be stored on a secure, regularly backed-up University network drive or computer within your organisational unit; and not
• • On your personal PC at home (though you can store a working copy of your data there)
  • On Dropbox or some other commercial cloud storage facility (not secure!)
  • On a USB stick – whether stored on campus, or carried around in your pocket / handbag for convenience (again, not secure!) …
• Data Retention: once the active phase of your project is complete, you must ensure data are moved to a secure repository until time to delete (especially important for Honours projects, where the student may well leave the University at the very start of the 5 year retention period)

This is a fairly common situation – many people find that data they have gathered as part of their teaching or workplace activities are sufficiently interesting to justify publishing in an academic venue. The issue is how to obtain approval to re-use these gathered data in the new way.

There are two possible approaches to applying for approval to use such data:

• If the data were gathered from identifiable sources (e.g. via interview) the simplest approach is to submit a minimal risk application, with an Information Sheet and Consent Form explaining your plans and asking the original participants for their approval to use the already-collected data in the new way;
• If, however, the data are non-identifiable, you will need to request a 'waiver of consent' under Section 2.3.6. of the National Statement

The National Statement handles waiver of consent in Section 2.3.6 by means of a series of questions the researcher/s need to answer explicitly in their HRE application. If you include this list of questions and your responses in a suitable part of the minimal risk application form (e.g. Section 9 - Recruitment of Participants or Section 15 – Consent Form) that, together with an explanation of your situation, is normally sufficient.

"Before deciding to waive the requirement for consent ... an HREC ... must be satisfied that:

1. involvement in the research carries no more than low risk ... to participants
2. the benefits from the research justify any risks of harm associated with not seeking consent
3. it is impracticable to obtain consent (e.g. due to the quantity, age or accessibility of records)
4. there is no known or likely reason for thinking that participants would not have consented if they had been asked
5. there is sufficient protection of their privacy
6. there is an adequate plan to protect the confidentiality of data
7. in case the results have significance for the participants' welfare there is, here practicable, a plan for making information arising from the research available to them (e.g. via a website)
8. the possibility of commercial exploitation of derivatives of the data ... will not deprive the participants of any financial benefits to which they would be entitled
9. the waiver is not prohibited by State, federal or international law"

Many researchers (especially those engaged in qualitative research) find it difficult to answer this question (which is found in Section 10 of the full application form). The following examples, adapted from previous ethics applications– both high and minimal risk – and completely anonymised (or, in one case, written by me entirely) may provide assistance. But note that you will need to chase up the references yourself (I haven't included them here). And the text included within <> symbols will need to be replaced with an equivalent phrase relating to your project …

Example 1 (justifying a relatively small sample of interviewees):Many authors on qualitative research methods agree that saturation is achieved at a comparatively low level (e.g. Guest et al., 2006), and samples generally do not need to be greater than 60 participants (Charmaz, 2006) with 15 as the smallest acceptable sample (adapted from Guest et al., 2006). Other researchers have provided similar guidelines for qualitative sample size, e.g.

• Charmaz (2006, p.114) suggests that "25 [participants are] adequate for smaller projects"
• Ritchie et al. (2003, p.84) note that qualitative samples often "lie under 50"; and
• Green & Thorogood (2009, p.120) state that "the experience of most qualitative researchers is that in interview studies little that is 'new' comes out of transcripts after you have interviewed 20 or so people".

Example 2 (Justifying a statistically analysable sample):A minimum of 30 participants would be required for the statistical analysis of the data to yield meaningful results. We would hope to recruit more than 100 participants, so that the sample size can be compared to those found in previous studies from different <equivalent strata> (e.g. Park & Chitiyo, 2011).

Example 3 (Justifying a questionnaire or other qualitative tool):To ensure a valid assessment of the proposed <instrument> it is estimated that the validation of the full <tool> will require a minimum of 200 participants (Tabachnick & Fidell, 2001). Statisticians recommend that in designing questionnaires and/or assessments at least 5-10 participants per variable are required – with an ideal sample size of 300 participants (Comrey & Lee, 1992; Field, 2005).

Example 4 (Justifying focus group size):Focus group size is not a well-researched topic: Carlsen & Glenton (2011) reviewed 220 papers published in 117 journals reporting on focus groups and found only 37 provided a justification for their group size, with a variety of relatively unscientific justifications – pragmatic approaches to focus group size justification being the most frequent. Gill, Stewart, Treasure & Chadwick (2008) cite the views of Stewart & Shamdasani (1990) and advise over-recruiting – and potentially managing a rather larger than optimal group – rather than under-recruiting, to avoid having to cancel the session or having an unsatisfactory discussion. These authors believe the optimum size for a focus group is 6-8 participants (excluding researchers), but suggest a workable range of 3-14 participants. Bloor, Frankland, Thomas & Robson (2001) note that small groups run the risk of limited discussion, while large groups can be chaotic, hard to manage for the facilitator and frustrating for participants who feel they get insufficient opportunities to speak.

Example 5 (Justifying experimental group size):We are proposing to collect 23 participants in each group based on a power analysis conducted in GPower3. [Unfortunately, we could not obtain an effect size from previous studies examining the <topic under research> as no previous studies have examined this question, however] we have based this power analysis on an effect size of differences in <dependent factor> in <experimental group> vs. <control group> (effect size .2: Peri et al., 2000) using a standardized <experimental task>. To obtain a power of .8, with an alpha value of .05, using a repeated measures mixed factorial design with 3 groups and 4 measurements (response to the CS+ and CS-, prior to and after the <experimental> phase), GPower reports that we need a total sample size of 69 (approximately 23 in each group).

If you are a student researcher, it's best to commence the section "Who is conducting the project?" in your Information Sheet by making it clear this research project is being conducted in partial fulfilment of a <degree name> for <student name> and is being supervised by <supervisor name/s and affiliation details>. Remember to include all your supervisors.

Yes. On each of the information sheets include the ethics committee contact details as follows:

This study has been approved by the Tasmanian Social Sciences Human Research Ethics Committee. If you have concerns or complaints about the conduct of this study, please contact the Executive Officer of the HREC (Tasmania) Network on (03) 6226 7479 or email human.ethics@utas.edu.au. The Executive Officer is the person nominated to receive complaints from research participants. Please quote ethics reference number H_____.